The human and mouse enteric nervous system at single cell resolution
Regev Lab - Cell 2020
true
Compiled: November 27, 2020
Abstract
“The enteric nervous system ENS coordinates diverse functions in the intestine, but has eluded comprehensive molecular characterization due to the rarity and diversity of cells Here, we develop two methods to profile the ENS of adult mice and humans at single cell resolution: RAISIN RNA-Seq, for profiling intact nuclei with ribosome-bound mRNA, and MIRACL-Seq, for label-free enrichment of rare cell types by droplet-based profiling The 1,187,535 nuclei in our mouse atlas include 5,068 neurons from the ileum and colon, revealing extraordinary neuron diversity We highlight circadian expression changes in enteric neurons, show that disease-related genes are dysregulated with aging, and identify differences between the ileum and proximal distal colon In human, we profile 436,202 nuclei recovering 1,445 neurons and identify conserved and species-specific transcriptional programs, and putative neuro-epithelial, neuro-stromal, and neuro-immune interactions The human ENS expresses risk genes for neuropathic, inflammatory, and extra-intestinal diseases, suggesting neuronal contributions to disease In this analysis the mouse mli Neuron and mli Glia subset is re-analysed”Data Preparation
Set Environment and load R
## Get interactive session ##
# srun --time=08:00:00 --mem=40G -p int --pty bash
# module purge;source /camp/stp/babs/working/software/modulepath_new_software_tree_2018-08-13;module load pandoc/2.2.3.2-foss-2016b;ml R/3.6.0-foss-2016b-BABS;R;
# sbatch --time=08:00:00 --wrap "module purge;source /camp/stp/babs/working/software/modulepath_new_software_tree_2018-08-13;module load pandoc/2.2.3.2-foss-2016b;ml R/3.6.0-foss-2016b-BABS;Rscript runB.r" --job-name="rB" --mem=42G -o rB.slurm >> commands.txt
# sbatch --time=18:00:00 --wrap "module purge;source /camp/stp/babs/working/software/modulepath_new_software_tree_2018-08-13;module load pandoc/2.2.3.2-foss-2016b;ml R/3.6.0-foss-2016b-BABS;Rscript runB.r" --job-name="rB" -p hmem --mem=300G -o rB.slurm >> commands.txt
Set Parameters
library(tidyverse)
library(dplyr)
library(Seurat)
library(ggplot2)
library(tidyr)
library(knitr)
## Setup plot collection object
VersionPdfExt <- paste0(".V", gsub("-", "", Sys.Date()), ".pdf")
if (dir.exists("/Volumes/babs/working/boeings/")){
hpc.mount <- "/Volumes/babs/working/boeings/"
} else if (dir.exists("Y:/working/boeings/")){
hpc.mount <- "Y:/working/boeings/"
} else if (dir.exists("/camp/stp/babs/working/boeings/")){
hpc.mount <- "/camp/stp/babs/working/boeings/"
} else {
hpc.mount <- ""
}
# source(
# paste0(
# hpc.mount,
# "Stefan/protocol_files/github/boeings/packages/packageSourceCode/SBwebtools.pckg.r"
# )
# )
# source(
# paste0(
# hpc.mount,
# "Stefan/protocol_files/github/boeings/packages/scTools/scTools.r"
# )
# )
source("assets/scTools.r")
source("assets/SBwebtools.pckg.r")
####
FN <- paste0(hpc.mount, "Projects/reference_data/documentation/BC.parameters.txt")
dbTable <- read.delim(
FN,
sep = "\t",
stringsAsFactors = F
)
db.pwd <- as.vector(dbTable[1,1])
if (length(.libPaths()) > 2){
.libPaths(.libPaths()[2:3])
}
ObioFN <- paste0("../", list.files("..")[grep(".bioLOGIC.Robj", list.files(".."))])
load(ObioFN)
# checkFile = paste0(
# Obio@parameterList$project_id,
# ".bioLOGIC.Robj"
# )
#
# if (ObioFN != checkFile){
# exit()
# }
Obio <- setMountingPoint(Obio)
Obio <- setAnalysisPaths(Obio)
Obio <- setCrickGenomeAndGeneNameTable(Obio)
Obio <- createAnalysisFolders(
Obio,
baseDir="/camp/stp/babs/working/boeings/Projects/",
localBaseDir = paste0(hpc.mount, "Projects/")
)
Obio <- setDataBaseParameters(Obio)
Obio <- addGeneAnnotation(Obio)
Obio@parameterList[["reportFigDir"]] <- paste0(Obio@parameterList$localWorkDir,Obio@parameterList$project_id, "/report_figures/")
## Create outputfolders ##
if (!dir.exists(paste0(Obio@parameterList$localWorkDir,Obio@parameterList$project_id))){
dir.create(paste0(Obio@parameterList$localWorkDir,Obio@parameterList$project_id))
}
if (!dir.exists(Obio@parameterList$reportFigDir)){
dir.create(Obio@parameterList$reportFigDir)
}
figureCount <- 1
## Load R module load R/3.5.1-foss-2018b ##
#setwd(Obio@parameterList$localWorkDir)
if (Obio@parameterList$host == "10.27.241.234"){
urlString <- "biologic.thecrick.org"
} else {
urlString <- "biologic.crick.ac.uk"
}###############################################################################
## Compile reference gene list ##
## Set gene reference list
Obio@dataTableList[["referenceList"]] <- list()
if (is.null(Obio@parameterList$HmDisplayCatsFromDb)){
Obio@parameterList$HmDisplayCatsFromDb <- list("TFs" = "ag_lab_categories__10")
}
## Add transcription factors that happen to be cluster markers ##
catList <- Obio@parameterList$HmDisplayCatsFromDb
if (Obio@parameterList$geneIDcolumn != "mgi_symbol" & Obio@parameterList$geneIDcolumn != "hgnc_symbol") {
queryGS <- "hgnc_symbol"
} else {
queryGS <- Obio@parameterList$geneIDcolumn
}
for (i in 1:length(catList)){
tempVec <- retrieve.gene.category.from.db(
cat_id = catList[[i]],
password = db.pwd,
gene.symbol = queryGS,
user = Obio@parameterList$db.user,
host = Obio@parameterList$host
)
Obio@dataTableList$referenceList[[names(catList)[i]]] <- tempVec
}
## Transcription Factors in variable genes ##
#tempVec <- tempVec[tempVec %in% Obio@dataTableList$dfGeneralMarkers$gene]Create Sample List
###############################################################################
## Create sample list filtered on MT and norm_counts_RNA ##
SampleList <- createNormSampleList(
obj = Obio,
reduce = NULL
)
## Done ##
###############################################################################
###############################################################################
## Add doublet annotation, if present, to meta data ##
pos <- grep("DF_resultlist", names(Obio@dataTableList))
if (length(pos) > 0){
sampleNames <- names(SampleList)
for (i in 1:length(SampleList)){
dfAdd <- Obio@dataTableList[["DF_resultlist"]][[sampleNames[i]]]
row.names(dfAdd) <- gsub("-1", "",row.names(dfAdd))
dfAdd <- dfAdd[row.names(dfAdd) %in% row.names(SampleList[[i]]@meta.data),]
SampleList[[i]] <- addDf2seuratMetaData(
obj = SampleList[[i]],
dfAdd = dfAdd
)
}
}
## Done ##
###############################################################################
###############################################################################
## Integrate Datasets ##
if (length(SampleList) > 1){
if (Obio@parameterList$scIntegrationMethod == "SCT"){
if (length(grep("scNintegrationFeatures", names(Obio@parameterList))) == 0){
Obio@parameterList$scNintegrationFeatures = 3000
}
library(future)
options(future.globals.maxSize = 14000 * 1024^2)
plan("multiprocess", workers = 30)
sample.features <- SelectIntegrationFeatures(
object.list = SampleList,
nfeatures = Obio@parameterList$scNintegrationFeatures
)
SampleList <- PrepSCTIntegration(
object.list = SampleList,
anchor.features = sample.features,
verbose = FALSE
)
sampleAnchors <- FindIntegrationAnchors(
object.list = SampleList,
normalization.method = "SCT",
anchor.features = sample.features,
verbose = FALSE
)
OsC <- IntegrateData(
anchorset = sampleAnchors,
normalization.method = "SCT",
verbose = FALSE
)
detach("package:future", unload=TRUE)
} else {
sampleAnchors <- FindIntegrationAnchors(
object.list = SampleList,
dims = 1:30
)
OsC <- IntegrateData(
#features.to.integrate = geneIntersectVec,
anchorset = sampleAnchors,
dims = 1:30
)
}
Obio@dataTableList$referenceList[["sampleAnchors"]] <- as.vector(sort(sampleAnchors@anchor.features))
} else {
OsC <- SampleList[[1]]
}
Idents(OsC) <- factor(Idents(OsC), levels = names(Obio@sampleDetailList))
dotsize = 1
if (nrow(OsC@meta.data) > 10000){
dotsize = 0.75
} else if (nrow(OsC@meta.data) > 20000){
dotsize = 0.5
} else if (nrow(OsC@meta.data) > 50000){
dotsize = 0.25
}
## Done integration ##
###############################################################################
###############################################################################
## Add additional annotation data ##
# ###############################################################################
# ## Load extra annotation data ##
# FN <- "/camp/stp/babs/working/boeings/Projects/swantonc/dhruva.biswas/335_CSL_DB_scRNA_seq_lung_tumor_microenvironment_EMTAB6653/basedata/MetaData.txt"
#
# dfExtra <- read.delim(
# FN,
# header = T,
# sep = "\t",
# stringsAsFactors = F
# )
#
# dfExtra <- dfExtra[dfExtra$cell %in% row.names(OsC@meta.data),]
#
# names(dfExtra) <- gsub("^cluster$", "Article_Cluster", names(dfExtra))
# names(dfExtra) <- gsub("^CellType$", "Article_Cell_Type", names(dfExtra))
# names(dfExtra) <- gsub("^CellType$", "Article_Cell_Type", names(dfExtra))
# names(dfExtra) <- gsub("^CellType$", "Article_Cell_Type", names(dfExtra))
#
# dfExtra[["Patient"]] <- sapply(dfExtra$Patient_piece, function(x) unlist(strsplit(x, "_"))[1])
# dfExtra$Patient <- paste0("P", dfExtra$Patient)
# dfExtra[["Region"]] <- sapply(dfExtra$Patient_piece, function(x) unlist(strsplit(x, "_"))[2])
# dfExtra$CellFromTumor <- as.character(dfExtra$CellFromTumor)
# dfExtra$CellFromTumor[dfExtra$CellFromTumor == "TRUE"] <- "Tumor"
# dfExtra$CellFromTumor[dfExtra$CellFromTumor == "FALSE"] <- "Non-Tumor"
#
#
#
# row.names(dfExtra) <- dfExtra$cell
#
# selVec <- c("CellFromTumor", "Patient", "Region")
# dfAdd <- dfExtra[,selVec]
#
# OsC <- addDf2seuratMetaData(
# obj = OsC,
# dfAdd = dfAdd
# )
#
# ## Done extra annotation ##
# ###############################################################################plotList <- list()
chnkVec <- as.vector(NULL, mode = "character")
## First make variation plot for integrated samples, than for all individual samples separately
tag <- "Integrated_Samples"
DefaultAssay(OsC) <- "RNA"
OsC <- FindVariableFeatures(
object = OsC,
selection.method = 'vst',
nfeatures = 2000
)
# Identify the 10 most highly variable genes
label2000 <- paste0("integrated", "_", "top2000var")
Obio@dataTableList$referenceList[[ label2000]]<- head(
x = VariableFeatures(object = OsC),
2000
)
label30 <- paste0("integrated", "_", "top30var")
Obio@dataTableList$referenceList[[ label30]]<- head(
x = VariableFeatures(object = OsC),
30
)
## slot for variable features OsC@assays$RNA@var.features
dfVar <- OsC@assays$RNA@meta.features
names(dfVar) <- gsub("vst.", "",names(dfVar))
dfVar[["gene"]] <- row.names(dfVar)
OsC@meta.data[["all"]] <- "all"
Idents(OsC) <- "all"
cluster.averages <- AverageExpression(
OsC,
return.seurat = TRUE
)
Idents(OsC) <- "sampleID"
dfAvgExpr <- data.frame(cluster.averages[["RNA"]]@data)
dfAvgExpr[["gene"]] <- row.names(dfAvgExpr)
names(dfAvgExpr)[1] <- "Avg.Expression"
dfVar <- merge(
dfVar,
dfAvgExpr,
by.x = "gene",
by.y = "gene"
)
dfVar[["Type"]] <- "Standard"
dfVar[dfVar$gene %in% OsC@assays$RNA@var.features, "Type"] <- "Most Variable"
dfVar[["text"]] <- ""
dfVar[dfVar$gene %in% as.vector(Obio@dataTableList$referenceList[[label30]]), "text"] <- Obio@dataTableList$referenceList[label30]
#dotsize <- 0.5
library(ggrepel)
plotList[[tag]] <- ggplot(
data = dfVar,
aes(
x=Avg.Expression,
y=variance.standardized, label = text, color = Type
)
) + geom_point( shape=16, size = dotsize
) + xlab("Average Expression") + ylab("Variance Standarized"
) + theme_bw(
) + theme(
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12)
) + ggtitle(paste0("Variance vs. Expression in the Overall Experiment")
) + scale_color_manual(values=c("#FF0000", "#000000")
) + geom_text_repel(
)
###########################################################################
## Save plot to file ##
FNbase <- paste0("variation.integrated.samples.", VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
## ##
###########################################################################
link <- paste0('<a href="https://',urlString,'/',Obio@parameterList$project_id,'/scatterplot?x_axis=',paste0(tag, '_AvgExpr'),'&y_axis=',paste0(tag, '_var_std" target = "_blank">here</a>'))
figCap <- paste0(
'**Figure ',
figureCount,
':** Variance versus averaged gene expression for overall sample.',
names(SampleList)[i],
'. ',
'Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>. ',
'An interactive version of this figure can be found ', link, '. '
)
NewChnk <- paste0(
"#### ",tag,
"\n```{r varplot_",tag,", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",figCap,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
## Histogram Part C done ##
###########################################################################
chnkVec <- c(
chnkVec,
NewChnk
)
figureCount <- figureCount + 1
print(" All variation done.")
## Now the individual samples ##
xmax <- ceiling(max(dfVar$Avg.Expression))
ymax <- ceiling(max(dfVar$variance.standardized))
dfVarRes <- unique(
dfVar[,c("gene", "Avg.Expression", "variance.standardized")]
)
names(dfVarRes) <- gsub(
"Avg.Expression", paste0(tag, "_AvgExpr"), names(dfVarRes)
)
names(dfVarRes) <- gsub(
"variance.standardized", paste0(tag, "_var_std"), names(dfVarRes)
)
for (i in 1:length(SampleList)){
tag <- paste0("Ind_var_", names(SampleList)[i])
DefaultAssay(SampleList[[i]]) <- "RNA"
SampleList[[i]] <- FindVariableFeatures(object = SampleList[[i]])
SampleList[[i]] <- FindVariableFeatures(
object = SampleList[[i]],
selection.method = 'vst',
nfeatures = 2000
)
# Identify the 10 most highly variable genes
label2000 <- paste0(names(SampleList)[i], "_", "top2000var")
Obio@dataTableList$referenceList[[ label2000]]<- head(
x = VariableFeatures(object = SampleList[[i]]),
2000
)
label30 <- paste0(names(SampleList)[i], "_", "top30var")
Obio@dataTableList$referenceList[[ label30]]<- head(
x = VariableFeatures(object = SampleList[[i]]),
30
)
## slot for variable features OsC@assays$RNA@var.features
dfVar <- SampleList[[i]]@assays$RNA@meta.features
names(dfVar) <- gsub("vst.", "",names(dfVar))
dfVar[["gene"]] <- row.names(dfVar)
cluster.averages <- AverageExpression(
SampleList[[i]],
return.seurat = TRUE
)
dfAvgExpr <- data.frame(cluster.averages[["RNA"]]@data)
dfAvgExpr[["gene"]] <- row.names(dfAvgExpr)
names(dfAvgExpr)[1] <- "Avg.Expression"
dfVar <- merge(
dfVar,
dfAvgExpr,
by.x = "gene",
by.y = "gene"
)
dfVar[["Type"]] <- "Standard"
dfVar[dfVar$gene %in% SampleList[[i]]@assays$RNA@var.features, "Type"] <- "Most Variable"
dfVar[["text"]] <- ""
dfVar[dfVar$gene %in% as.vector(Obio@dataTableList$referenceList[[label30]]), "text"] <- Obio@dataTableList$referenceList[label30]
# dotsize <- 0.5
library(ggrepel)
plotList[[tag]] <- ggplot(
data = dfVar,
aes(
x=Avg.Expression,
y=variance.standardized, label = text, color = Type)
) + geom_point( shape=16, size = dotsize
) + xlab("Average Expression") + ylab("Variance Standarized"
) + theme_bw(
) + theme(
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12)
)+ ggtitle(paste0("Individual variance in the ", names(SampleList)[i], " sample.")
) + scale_color_manual(values=c("#FF0000", "#000000")
) + geom_text_repel(
) + xlim(0, xmax) + ylim(0, ymax
)
###########################################################################
## Save plot to file ##
FNbase <- paste0("Individual.var.features",names(SampleList)[i], VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
## ##
###########################################################################
link <- paste0('<a href="https://',urlString,'/',Obio@parameterList$project_id,'/scatterplot?x_axis=',paste0(tag, '_AvgExpr'),'&y_axis=',paste0(tag, '_var_std" target="_blank">here</a>'))
figCap <- paste0(
'**Figure ',
figureCount,
':** Variance versus averaged gene expression for sample ',
names(SampleList)[i],
'. ',
'Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>. ',
'An interactive version of this figure can be found ', link, '. '
)
NewChnk <- paste0(
"#### ",tag,
"\n```{r varplot_",tag,", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",figCap,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
## Histogram Part C done ##
###########################################################################
chnkVec <- c(
chnkVec,
NewChnk
)
figureCount <- figureCount + 1
print(paste0(names(SampleList)[i], " individual variation done."))
## Add to result table ##
dfVarTemp <- unique(
dfVar[,c("gene", "Avg.Expression", "variance.standardized")]
)
names(dfVarTemp) <- gsub(
"Avg.Expression", paste0(tag, "_AvgExpr"), names(dfVarTemp)
)
names(dfVarTemp) <- gsub(
"variance.standardized", paste0(tag, "_var_std"), names(dfVarTemp)
)
dfVarRes <- merge(
dfVarRes,
dfVarTemp,
by.x = "gene",
by.y = "gene",
all = TRUE
)
dfVarRes[is.na(dfVarRes)] <- 0
}
## Make sure summary goes first ##
Obio@dataTableList[["dfVariation"]] <- dfVarRes
if (length(plotList) > 3){
tabVar <- ".tabset .tabset-fade .tabset-dropdown"
} else {
tabVar <- ".tabset .tabset-fade .tabset-pills"
}Result Figures
Quality Control Plots: Variable Features in the combined and all individual dataset
## plot list will be integrated in full figure ##
cat(paste(knit(text = chnkVec, quiet = T), collapse = '\n'))Integrated_Samples
Ind_var_mliGlia
Ind_var_mliNeuron
Figure 3: Variance versus averaged gene expression for sample mliNeuron. Download a pdf of this figure here. An interactive version of this figure can be found here.
###############################################################################
## Perform integrated analysis ##
if (length(Obio@sampleDetailList) > 1){
DefaultAssay(OsC) <- "integrated"
} else {
Obio@parameterList$singleCellClusterString <- gsub("integrated", "RNA", Obio@parameterList$singleCellClusterString)
}
# Run the standard workflow for visualization and clustering
## This will scale on the most variable features only
OsC <- ScaleData(OsC, verbose = FALSE)
OsC <- RunPCA(
OsC,
npcs = Obio@parameterList$singleCellSeuratNpcs4PCA, verbose = FALSE
)
# t-SNE and Clustering
## Add PCA clusters to data collection ##
OsC <- RunUMAP(OsC, reduction = "pca", dims = 1:20)
OsC <- RunTSNE(OsC, reduction = "pca", dims = 1:20)
OsC <- FindNeighbors(OsC, reduction = "pca", dims = 1:20)
OsC <- FindClusters(OsC, resolution = Obio@parameterList$singleCellClusterParameter)
## Rational: Run PCA on variable features, then scale data for heatmaps and other applications
if (Obio@parameterList$scIntegrationMethod != "RNA"){
DefaultAssay(OsC) <- Obio@parameterList$scIntegrationMethod
allGenes <- rownames(x = OsC@assays[[Obio@parameterList$scIntegrationMethod]])
OsC <- ScaleData(OsC, verbose = FALSE, features=allGenes)
}
DefaultAssay(OsC) <- "RNA"
allGenes <- rownames(x = OsC@assays$RNA)
OsC <- ScaleData(OsC, verbose = FALSE, features=allGenes)If you wish to get a bit of background on tSNE dimensionality reduction, take a look at this youtube video by Josh Starmer from the University of North Carolina.
If you wish to get a bit of background on UMAP (and other) dimensionality reduction algorithms, take a look at this youtube video by recaping a lecture at the PyData 2018 conference.
reductionVec <- c("umap", "tsne")
plotList <- list()
chnkVec <- as.vector(NULL, mode = "character")
###############################################################################
## First UMAP all samples together ##
tag <- paste0("UMAP_sample_level")
dfPlot <- OsC@meta.data
pos <- grep("included", names(dfPlot))
if (length(pos) == 0){
dfPlot[["included"]] <- "+"
}
pos <- grep("cellID", names(dfPlot))
if (length(pos) == 0){
dfPlot[["cellID"]] <- row.names(dfPlot)
}
dfPlot$UMAP_1 <- NULL
dfPlot$UMAP_2 <- NULL
## Get UMAP coordinates ##
coord <- data.frame(OsC@reductions$umap@cell.embeddings)
coord[["cellID"]] <- row.names(coord)
coord <-coord[coord$cellID %in% dfPlot$cellID, ]
dfPlot <- merge(dfPlot, coord, by.x = "cellID", by.y="cellID", all=T)
dfPlot[is.na(dfPlot)] <- 0
dfPlot <- dfPlot[dfPlot$UMAP_1 != 0 & dfPlot$UMAP_2 != 0,]
## Add cluster colors ##
dfPlot[["Cluster"]] <- dfPlot$sampleID
clusterVec <- as.vector(sort(unique(dfPlot$sampleID)))
maxX <- 1.1*max(dfPlot$UMAP_1, na.rm = T)
minX <- 1.1*min(dfPlot$UMAP_1, na.rm = T)
maxY <- 1.1*max(dfPlot$UMAP_2, na.rm = T)
minY <- 1.1*min(dfPlot$UMAP_2, na.rm = T)
library(scales)
clusterCols = hue_pal()(length(clusterVec))
dfPlot$Cluster <- factor(dfPlot$Cluster, levels = clusterVec)
# dotsize = 1
# if (nrow(dfPlot) > 10000){
# dotsize = 0.75
# } else if (nrow(dfPlot) > 20000){
# dotsize = 0.5
# } else if (nrow(dfPlot) > 50000){
# dotsize = 0.25
# }
plotList[[tag]] <- ggplot(data=dfPlot[dfPlot$included == "+",], aes(UMAP_1, UMAP_2, color=sampleID)
) + geom_point( shape=16, size = as.numeric(dotsize)
) + xlab("UMAP1") + ylab("UMAP2"
) + theme_bw(
) + theme(
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12),
legend.title = element_blank()
) + ggtitle(paste0("Sample: ", tag)
) + xlim(minX, maxX) + ylim(minY, maxY
) + coord_fixed(ratio=1
)
if (length(unique(dfPlot$Cluster)) > 15){
plotList[[tag]] <- plotList[[tag]] + theme(legend.position = "none")
}
FNbase <- paste0(tag, VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
link <- paste0('<a href="https://',urlString,'/',Obio@parameterList$project_id,'/pca?x_axis=UMAP_1&y_axis=UMAP_2" target="_blank">here</a>')
figLegend <- paste0(
'**Figure ',
figureCount,
':** ',
' UMAP showing all cells from all samples together. Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>.',
'An interactive version of this figure can be found ', link, '. '
)
figureCount <- figureCount + 1
NewChnk <- paste0(
"#### ", tag,
"\n```{r Sample_UMAP_",
tag,", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
## Done first umap all samples ##
###############################################################################
###############################################################################
## First tsne all samples together ##
tag <- paste0("tSNE_All_Samples")
dfPlot <- OsC@meta.data
pos <- grep("included", names(dfPlot))
if (length(pos) == 0){
dfPlot[["included"]] <- "+"
}
pos <- grep("cellID", names(dfPlot))
if (length(pos) == 0){
dfPlot[["cellID"]] <- row.names(dfPlot)
}
dfPlot$tSNE_1 <- NULL
dfPlot$tSNE_2 <- NULL
## Get tsNE coordinates ##
coord <- data.frame(OsC@reductions$tsne@cell.embeddings)
coord[["cellID"]] <- row.names(coord)
coord <-coord[coord$cellID %in% dfPlot$cellID, ]
dfPlot <- merge(dfPlot, coord, by.x = "cellID", by.y="cellID", all=T)
dfPlot[is.na(dfPlot)] <- 0
dfPlot <- dfPlot[dfPlot$tSNE_1 != 0 & dfPlot$tSNE_2 != 0,]
## Add cluster colors ##
dfPlot[["Cluster"]] <- dfPlot$sampleID
clusterVec <- as.vector(unique(sort(dfPlot$sampleID)))
maxX <- 1.1*max(dfPlot$tSNE_1, na.rm = T)
minX <- 1.1*min(dfPlot$tSNE_1, na.rm = T)
maxY <- 1.1*max(dfPlot$tSNE_2, na.rm = T)
minY <- 1.1*min(dfPlot$tSNE_2, na.rm = T)
library(scales)
clusterCols = hue_pal()(length(clusterVec))
dfPlot$Cluster <- factor(dfPlot$Cluster, levels = clusterVec)
# dotsize = 1.5
# if (nrow(dfPlot) > 10000){
# dotsize = 0.75
# } else if (nrow(dfPlot) > 50000){
# dotsize = 0.5
# } else {
# dotsize = 0.25
# }
plotList[[tag]] <- ggplot(data=dfPlot[dfPlot$included == "+",], aes(tSNE_1, tSNE_2, color=sampleID)
) + geom_point( shape=16, size = as.numeric(dotsize)
) + xlab("tSNE1") + ylab("tSNE2"
) + theme_bw(
) + theme(
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12),
legend.title = element_blank()
) + ggtitle(paste0("Sample: ", tag)
) + xlim(minX, maxX) + ylim(minY, maxY
) + coord_fixed(ratio=1
)
if (length(unique(dfPlot$Cluster)) > 15){
plotList[[tag]] <- plotList[[tag]] + theme(legend.position = "none")
}
FNbase <- paste0(tag, VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
link <- paste0('<a href="https://',urlString,'/',Obio@parameterList$project_id,'/pca?x_axis=tSNE_1&y_axis=tSNE_2" target="_blank">here</a>')
figLegend <- paste0(
'**Figure ',
figureCount,
':** ',
' tSNE showing all cells from all samples together. Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>.',
'An interactive version of this figure can be found ', link, '. '
)
figureCount <- figureCount + 1
NewChnk <- paste0(
"#### ", tag,
"\n```{r Sample_tSNE_",
tag,", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
## Done first tsne all samples ##
###############################################################################
###############################################################################
## Make one UMAP plot per sample ##
sampleVec <- sort(unique(OsC@meta.data$sampleID))
dfPlot <- OsC@meta.data
pos <- grep("included", names(dfPlot))
if (length(pos) == 0){
dfPlot[["included"]] <- "+"
}
pos <- grep("cellID", names(dfPlot))
if (length(pos) == 0){
dfPlot[["cellID"]] <- row.names(dfPlot)
}
## Get UMAP coordinates ##
coord <- data.frame(OsC@reductions$umap@cell.embeddings)
coord[["cellID"]] <- row.names(coord)
coord <-coord[coord$cellID %in% dfPlot$cellID, ]
dfPlot$UMAP_1 <- NULL
dfPlot$UMAP_2 <- NULL
dfPlot <- merge(dfPlot, coord, by.x = "cellID", by.y="cellID", all=T)
dfPlot[is.na(dfPlot)] <- 0
dfPlot <- dfPlot[dfPlot$UMAP_1 != 0 & dfPlot$UMAP_2 != 0,]
## Add cluster colors ##
dfPlot[["Cluster"]] <- dfPlot$sampleID
clusterVec <- as.vector(unique(sort(dfPlot$sampleID)))
library(scales)
clusterCols = hue_pal()(length(clusterVec))
dfPlot$Cluster <- factor(dfPlot$Cluster, levels = clusterVec)
maxX <- 1.1*max(dfPlot$UMAP_1, na.rm = T)
minX <- 1.1*min(dfPlot$UMAP_1, na.rm = T)
maxY <- 1.1*max(dfPlot$UMAP_2, na.rm = T)
minY <- 1.1*min(dfPlot$UMAP_2, na.rm = T)
for (i in 1:length(sampleVec)){
tag <- paste0("UMAP_plot_by_", sampleVec[i])
dfPlotSel <- dfPlot[dfPlot$sampleID == sampleVec[i], ]
plotList[[tag]] <- ggplot(data=dfPlotSel[dfPlot$included == "+",], aes(UMAP_1, UMAP_2, color=sampleID)
) + geom_point( shape=16, size = as.numeric(dotsize)
) + xlab("UMAP1") + ylab("UMAP2"
) + theme_bw(
) + theme(
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12),
legend.title = element_blank()
) + ggtitle(paste0("Sample: ", tag)
) + xlim(minX, maxX) + ylim(minY, maxY
) + coord_fixed(ratio=1
) + scale_color_manual(values = clusterCols[i]
)
if (length(unique(dfPlot$Cluster)) > 15){
plotList[[tag]] <- plotList[[tag]] + theme(legend.position = "none")
}
FNbase <- paste0(tag, VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
figLegend <- paste0(
'**Figure ',
figureCount,
':** ',
' Sample-level UMAPs. Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>.'
)
figureCount <- figureCount + 1
NewChnk <- paste0(
paste("#### ", tag),
"\n```{r Sample_UMAP_",
tag,", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
}
## Done making one umap plot per sample ##
###############################################################################
###############################################################################
## Add cluster dendrogram by sample ##
if (length(unique(OsC@meta.data$sampleID)) > 3){
library(ggtree)
Idents(OsC) <- "sampleID"
OsC <- BuildClusterTree(OsC)
tag <- paste0("Sample_Dendrogram")
OsC@tools$BuildClusterTree$tip.label <- paste0("C", OsC@tools$BuildClusterTree$tip.label)
plotList[[tag]] <- ggplot(OsC@tools$BuildClusterTree
) + geom_tree(
) + theme_tree(
) + geom_tiplab(
) + labs(title=tag
) + theme(
panel.border = element_rect(colour = "black", fill=NA, size=1),
axis.title.x=element_blank(),
plot.title = element_text(hjust = 0.5, size = 12)
) + xlim(0,1.2*max(OsC@tools$BuildClusterTree[[2]]))
## Save to file ##
FNbase <- paste0(tag,".", VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
figLegend <- paste0(
'**Figure ',
figureCount,
':** ',
' Clusterplot dendrogram by sample ID. ','A pdf of this figure can be downloaded <a href="',FNrel,'" target="_blank">here</a>.'
)
NewChnk <- paste0(
"#### SampleID Dendrogram",
"\n```{r ", tag, "results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
figureCount <- figureCount + 1
}
## Done by sample ##
###############################################################################
if (length(plotList) > 3){
tabVar <- ".tabset .tabset-fade .tabset-dropdown"
} else {
tabVar <- ".tabset .tabset-fade .tabset-pills"
}Dimensionality Reduction Plots by Sample
cat(paste(knit(text = chnkVec, quiet = T), collapse = '\n'))UMAP_sample_level
tSNE_All_Samples
UMAP_plot_by_mliGlia
Figure 6: Sample-level UMAPs. Download a pdf of this figure here.
UMAP_plot_by_mliNeuron
Figure 7: Sample-level UMAPs. Download a pdf of this figure here.
Calculate Average Expression Per Cluster
## Ensure that hmIdent exists ##
## Want to average by cluster and subset by sampleID ##
###############################################################################
## Calculataing average expression across all samples ##
OsC@meta.data[["all"]] <- "all"
Idents(OsC) <- "all"
cluster.averages <- AverageExpression(OsC, return.seurat = TRUE)
## Retrieved Scaled data ##
dfAvgExpr <- data.frame(cluster.averages[["RNA"]]@data)
#dfAvgExpr <- dfAvgExpr[,sort(names(dfAvgExpr))]
dfAvgExpr[["gene"]] <- row.names(dfAvgExpr)
dfAvgScaledData <- data.frame(cluster.averages[["RNA"]]@scale.data)
#dfAvgScaledData <- dfAvgScaledData[,sort(names(dfAvgScaledData))]
dfAvgScaledData[["gene"]] <- row.names(dfAvgScaledData)
Obio@dataTableList[["dfAvglg10ExprAll"]] <- dfAvgExpr
Obio@dataTableList[["dfAvglg10ExprAllScaled"]] <- dfAvgScaledData
## Done ##
###############################################################################
OsC@meta.data[["clustIdent"]] <- paste0(
"C_", OsC@meta.data[,Obio@parameterList$singleCellClusterString]
)
Idents(OsC) <- "clustIdent"
if (length(grep("sampleID", names(OsC@meta.data))) > 0){
cluster.averages <- AverageExpression(OsC, return.seurat = TRUE, add.ident = "sampleID")
} else {
cluster.averages <- AverageExpression(OsC, return.seurat = TRUE)
}
Idents(OsC) <- "sampleID"
## Retrieved Scaled data ##
dfAvgExpr <- data.frame(cluster.averages[["RNA"]]@data)
dfAvgExpr <- dfAvgExpr[,sort(names(dfAvgExpr))]
dfAvgExpr[["gene"]] <- row.names(dfAvgExpr)
dfAvgScaledData <- data.frame(cluster.averages[["RNA"]]@scale.data)
dfAvgScaledData <- dfAvgScaledData[,sort(names(dfAvgScaledData))]
dfAvgScaledData[["gene"]] <- row.names(dfAvgScaledData)
Obio@dataTableList[["dfAvglg10ExprByClusterBySample"]] <- dfAvgExpr
Obio@dataTableList[["dfAvglg10ExprByClusterBySampleScaled"]] <- dfAvgScaledData
###############################################################################
###############################################################################
## Average by Cluster ##
Idents(OsC) <- "clustIdent"
Idents(OsC) <- factor(Idents(OsC), levels = sort(levels(OsC)))
cluster.averages <- AverageExpression(OsC, return.seurat = TRUE)
dfAvgExpr <- data.frame(cluster.averages[["RNA"]]@data)
dfAvgExpr <- dfAvgExpr[,sort(names(dfAvgExpr))]
dfAvgExpr[["gene"]] <- row.names(dfAvgExpr)
dfAvgScaledData <- data.frame(cluster.averages[["RNA"]]@scale.data)
dfAvgScaledData <- dfAvgScaledData[,sort(names(dfAvgScaledData))]
dfAvgScaledData[["gene"]] <- row.names(dfAvgScaledData)
Obio@dataTableList[["dfAvglg10ExprPerCluster"]] <- dfAvgExpr
Obio@dataTableList[["dfAvglg10ExprPerClusterScaled"]] <- dfAvgScaledData
## Done Average by Cluster ##
###############################################################################
###############################################################################
## Average gene expression by sample ##
Idents(OsC) <- "sampleID"
cluster.averages <- AverageExpression(OsC, return.seurat = TRUE)
## Retrieved Scaled data ##
dfAvgExpr <- data.frame(cluster.averages[["RNA"]]@data)
selVec <- names(dfAvgExpr)
dfAvgExpr[["gene"]] <- row.names(dfAvgExpr)
selVec <- c("gene", selVec)
dfAvgExpr <- dfAvgExpr[,selVec]
dfAvgScaledData <- data.frame(cluster.averages[["RNA"]]@scale.data)
selVec <- names(dfAvgScaledData)
dfAvgScaledData[["gene"]] <- row.names(dfAvgScaledData)
selVec <- c("gene", selVec)
dfAvgScaledData <- dfAvgScaledData[,selVec]
Obio@dataTableList[["dfAvglg10ExprBySample"]] <- dfAvgExpr
Obio@dataTableList[["dfAvglg10ExprBySampleScaled"]] <- dfAvgScaledData
## Done average gene expression by sample ##
###############################################################################reductionVec <- c("umap", "tsne")
plotList <- list()
chnkVec <- as.vector(NULL, mode = "character")
###############################################################################
## First UMAP all samples together ##
tag <- paste0("UMAP_All_Samples")
dfPlot <- OsC@meta.data
pos <- grep("included", names(dfPlot))
if (length(pos) == 0){
dfPlot[["included"]] <- "+"
}
dfPlot[["cellID"]] <- row.names(dfPlot)
dfPlot$UMAP_1 <- NULL
dfPlot$UMAP_2 <- NULL
## Get UMAP coordinates ##
coord <- data.frame(OsC@reductions$umap@cell.embeddings)
coord[["cellID"]] <- row.names(coord)
coord <-coord[coord$cellID %in% dfPlot$cellID, ]
dfPlot <- merge(dfPlot, coord, by.x = "cellID", by.y="cellID", all=T)
dfPlot[is.na(dfPlot)] <- 0
dfPlot <- dfPlot[dfPlot$UMAP_1 != 0 & dfPlot$UMAP_2 != 0,]
## Add cluster colors ##
dfPlot[["Cluster"]] <- paste0("C", dfPlot$seurat_clusters)
clusterVec <- as.vector(paste0("C", unique(sort(dfPlot$seurat_clusters))))
maxX <- 1.1*max(dfPlot$UMAP_1, na.rm = T)
minX <- 1.1*min(dfPlot$UMAP_1, na.rm = T)
maxY <- 1.1*max(dfPlot$UMAP_2, na.rm = T)
minY <- 1.1*min(dfPlot$UMAP_2, na.rm = T)
library(scales)
clusterCols = hue_pal()(length(clusterVec))
dfPlot$Cluster <- factor(dfPlot$Cluster, levels = clusterVec)
# dotsize = 1
# if (nrow(dfPlot) > 10000){
# dotsize = 0.75
# } else if (nrow(dfPlot) > 20000){
# dotsize = 0.5
# } else if (nrow(dfPlot) > 50000){
# dotsize = 0.25
# }
plotList[[tag]] <- ggplot(data=dfPlot[dfPlot$included == "+",], aes(UMAP_1, UMAP_2, color=Cluster)
) + geom_point( shape=16, size = as.numeric(dotsize)
) + xlab("UMAP1") + ylab("UMAP2"
) + theme_bw(
) + theme(
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12),
legend.title = element_blank()
) + ggtitle(paste0("Sample: ", tag)
) + xlim(minX, maxX) + ylim(minY, maxY
) + coord_fixed(ratio=1
)
if (length(unique(dfPlot$Cluster)) > 15){
plotList[[tag]] <- plotList[[tag]] + theme(legend.position = "none")
}
FNbase <- paste0(tag, VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
link <- paste0('<a href="https://',urlString,'/',Obio@parameterList$project_id,'/pca?x_axis=UMAP_1&y_axis=UMAP_2" target="_blank">here</a>')
figLegend <- paste0(
'**Figure ',
figureCount,
':** ',
' UMAP showing all cells from all samples together. Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>.',
'An interactive version of this figure can be found ', link, '. '
)
figureCount <- figureCount + 1
NewChnk <- paste0(
"#### ", tag,
"\n```{r SL_UMAP_",
tag,", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
## Done first umap all samples ##
###############################################################################
###############################################################################
## First tsne all samples together ##
tag <- paste0("tSNE_All_Samples")
dfPlot <- OsC@meta.data
pos <- grep("included", names(dfPlot))
if (length(pos) == 0){
dfPlot[["included"]] <- "+"
}
dfPlot[["cellID"]] <- row.names(dfPlot)
dfPlot$tSNE_1 <- NULL
dfPlot$tSNE_2 <- NULL
## Get tsNE coordinates ##
coord <- data.frame(OsC@reductions$tsne@cell.embeddings)
coord[["cellID"]] <- row.names(coord)
coord <-coord[coord$cellID %in% dfPlot$cellID, ]
dfPlot <- merge(dfPlot, coord, by.x = "cellID", by.y="cellID", all=T)
dfPlot[is.na(dfPlot)] <- 0
dfPlot <- dfPlot[dfPlot$tSNE_1 != 0 & dfPlot$tSNE_2 != 0,]
## Add cluster colors ##
dfPlot[["Cluster"]] <- paste0("C", dfPlot$seurat_clusters)
clusterVec <- as.vector(paste0("C", unique(sort(dfPlot$seurat_clusters))))
maxX <- 1.1*max(dfPlot$tSNE_1, na.rm = T)
minX <- 1.1*min(dfPlot$tSNE_1, na.rm = T)
maxY <- 1.1*max(dfPlot$tSNE_2, na.rm = T)
minY <- 1.1*min(dfPlot$tSNE_2, na.rm = T)
library(scales)
clusterCols = hue_pal()(length(clusterVec))
dfPlot$Cluster <- factor(dfPlot$Cluster, levels = clusterVec)
# dotsize = 1.5
# if (nrow(dfPlot) > 10000){
# dotsize = 0.75
# } else if (nrow(dfPlot) > 50000){
# dotsize = 0.5
# } else {
# dotsize = 0.25
# }
plotList[[tag]] <- ggplot(data=dfPlot[dfPlot$included == "+",], aes(tSNE_1, tSNE_2, color=Cluster)
) + geom_point( shape=16, size = as.numeric(dotsize)
) + xlab("tSNE1") + ylab("tSNE2"
) + theme_bw(
) + theme(
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12),
legend.title = element_blank()
) + ggtitle(paste0("Sample: ", tag)
) + xlim(minX, maxX) + ylim(minY, maxY
) + coord_fixed(ratio=1
)
if (length(unique(dfPlot$Cluster)) > 15){
plotList[[tag]] <- plotList[[tag]] + theme(legend.position = "none")
}
FNbase <- paste0(tag, VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
link <- paste0('<a href="https://',urlString,'/',Obio@parameterList$project_id,'/pca?x_axis=tSNE_1&y_axis=tSNE_2" target="_blank">here</a>')
figLegend <- paste0(
'**Figure ',
figureCount,
':** ',
' tSNE showing all cells from all samples together. Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>.',
'An interactive version of this figure can be found ', link, '. '
)
figureCount <- figureCount + 1
NewChnk <- paste0(
"#### ", tag,
"\n```{r SL_tSNE_",
tag,", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
## Done first tsne all samples ##
###############################################################################
###############################################################################
## Make one UMAP plot per sample ##
sampleVec <- sort(unique(OsC@meta.data$sampleID))
dfPlot <- OsC@meta.data
pos <- grep("included", names(dfPlot))
if (length(pos) == 0){
dfPlot[["included"]] <- "+"
}
dfPlot[["cellID"]] <- row.names(dfPlot)
## Get UMAP coordinates ##
coord <- data.frame(OsC@reductions$umap@cell.embeddings)
coord[["cellID"]] <- row.names(coord)
coord <-coord[coord$cellID %in% dfPlot$cellID, ]
dfPlot$UMAP_1 <- NULL
dfPlot$UMAP_2 <- NULL
dfPlot <- merge(dfPlot, coord, by.x = "cellID", by.y="cellID", all=T)
dfPlot[is.na(dfPlot)] <- 0
dfPlot <- dfPlot[dfPlot$UMAP_1 != 0 & dfPlot$UMAP_2 != 0,]
## Add cluster colors ##
dfPlot[["Cluster"]] <- paste0("C", dfPlot$seurat_clusters)
clusterVec <- as.vector(paste0("C", unique(sort(dfPlot$seurat_clusters))))
library(scales)
clusterCols = hue_pal()(length(clusterVec))
dfPlot$Cluster <- factor(dfPlot$Cluster, levels = clusterVec)
maxX <- 1.1*max(dfPlot$UMAP_1, na.rm = T)
minX <- 1.1*min(dfPlot$UMAP_1, na.rm = T)
maxY <- 1.1*max(dfPlot$UMAP_2, na.rm = T)
minY <- 1.1*min(dfPlot$UMAP_2, na.rm = T)
for (i in 1:length(sampleVec)){
tag <- paste0("UMAP_plot_by_", sampleVec[i])
dfPlotSel <- dfPlot[dfPlot$sampleID == sampleVec[i], ]
plotList[[tag]] <- ggplot(data=dfPlotSel[dfPlot$included == "+",], aes(UMAP_1, UMAP_2, color=Cluster)
) + geom_point( shape=16, size = as.numeric(dotsize)
) + xlab("UMAP1") + ylab("UMAP2"
) + theme_bw(
) + theme(
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12),
legend.title = element_blank()
) + ggtitle(paste0("Sample: ", tag)
) + xlim(minX, maxX) + ylim(minY, maxY
) + coord_fixed(ratio=1
)
if (length(unique(dfPlot$Cluster)) > 15){
plotList[[tag]] <- plotList[[tag]] + theme(legend.position = "none")
}
FNbase <- paste0(tag, VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
figLegend <- paste0(
'**Figure ',
figureCount,
':** ',
' Sample-level UMAPs. Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>.'
)
figureCount <- figureCount + 1
NewChnk <- paste0(
paste("#### ", tag),
"\n```{r SL_UMAP_",
tag,", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
}
## Done making one umap plot per sample ##
###############################################################################
###############################################################################
## Add cluster dendrogram by sample ##
if (length(unique(OsC@meta.data$sampleID)) > 3){
library(ggtree)
Idents(OsC) <- "sampleID"
OsC <- BuildClusterTree(OsC)
tag <- paste0("Sample_Dendrogram")
OsC@tools$BuildClusterTree$tip.label <- paste0("C", OsC@tools$BuildClusterTree$tip.label)
plotList[[tag]] <- ggplot(OsC@tools$BuildClusterTree
) + geom_tree(
) + theme_tree(
) + geom_tiplab(
) + labs(title=tag
) + theme(
panel.border = element_rect(colour = "black", fill=NA, size=1),
axis.title.x=element_blank(),
plot.title = element_text(hjust = 0.5, size = 12)
) + xlim(0,1.2*max(OsC@tools$BuildClusterTree[[2]]))
## Save to file ##
FNbase <- paste0(tag,".", VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
figLegend <- paste0(
'**Figure ',
figureCount,
':** ',
' Clusterplot dendrogram by sample ID. ','A pdf of this figure can be downloaded <a href="',FNrel,'" target="_blank">here</a>.'
)
NewChnk <- paste0(
"#### SampleID Dendrogram",
"\n```{r ", tag, "results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
figureCount <- figureCount + 1
}
## Done by sample ##
###############################################################################
###############################################################################
## Add cluster dendrogram by cluster ##
library(ggtree)
Idents(OsC) <- "seurat_clusters"
OsC <- BuildClusterTree(OsC)
tag <- paste0("Cluster_Dendrogram")
OsC@tools$BuildClusterTree$tip.label <- paste0("C", OsC@tools$BuildClusterTree$tip.label)
plotList[[tag]] <- ggplot(OsC@tools$BuildClusterTree
) + geom_tree(
) + theme_tree(
) + geom_tiplab(
) + labs(title=tag
) + theme(
panel.border = element_rect(colour = "black", fill=NA, size=1),
axis.title.x=element_blank(),
plot.title = element_text(hjust = 0.5, size = 12)
) + xlim(0, 2*max(OsC@tools$BuildClusterTree[[2]]))
#+ xlim(-1,1.2*max(OsC@tools$BuildClusterTree$edge))
## Save to file ##
FNbase <- paste0(tag,".", VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
figLegend <- paste0(
'**Figure ',
figureCount,
':** ',
' Clusterplot dendrogram. ','A pdf of this figure can be downloaded <a href="',FNrel,'" target="_blank">here</a>.'
)
NewChnk <- paste0(
"#### Cluster Dendrogram",
"\n```{r ", tag, "results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
figureCount <- figureCount + 1
## Done integraed analysis ##
###############################################################################
###############################################################################
## Find all markers ##
DefaultAssay(OsC) <- "RNA"
Idents(OsC) <- Obio@parameterList$singleCellClusterString
lgFCthreshold <- 0.25
dfGeneralMarkers <- FindAllMarkers(
object = OsC,
only.pos = FALSE,
min.pct = 0.1,
logfc.threshold = lgFCthreshold,
test.use = "roc",
assay = "RNA",
slot = "data"
)
if (nrow(dfGeneralMarkers) == 0){
lgFCthreshold <- 0.00
dfGeneralMarkers <- FindAllMarkers(
object = OsC,
only.pos = FALSE,
min.pct = 0.01,
logfc.threshold = lgFCthreshold,
test.use = "roc",
assay = "RNA",
slot = "data"
)
}
if (nrow(dfGeneralMarkers) > 0){
dfGeneralMarkers[["direction"]] <- ""
dfGeneralMarkers[dfGeneralMarkers$avg_diff >= 0, "direction"] <- "positive"
dfGeneralMarkers[dfGeneralMarkers$avg_diff < 0, "direction"] <- "negative"
Obio@dataTableList[["dfGeneralMarkers"]] <- dfGeneralMarkers
dfGeneralMarkersFilt <- dfGeneralMarkers[dfGeneralMarkers$avg_diff > lgFCthreshold | dfGeneralMarkers$avg_diff < -lgFCthreshold,]
Obio@dataTableList[["dfGeneralMarkersFilt"]] <- dfGeneralMarkersFilt
dfTop1 <- data.frame(dfGeneralMarkers %>% group_by(cluster) %>% top_n(1, avg_diff))
dfTop5 <- data.frame(dfGeneralMarkers %>% group_by(cluster) %>% top_n(5, avg_diff))
dfTop10 <- data.frame(dfGeneralMarkers %>% group_by(cluster) %>% top_n(10, avg_diff))
Obio@dataTableList[["dfGeneralMarkersTop10"]] <- dfTop10
Obio@dataTableList$referenceList[["Top10clusterMarkers"]] <- as.vector(
unique(
dfTop10$gene
)
)
} else {
Obio@dataTableList[["dfGeneralMarkers"]] <- NULL
}
#############################################################
###############################################################################
## Upload general markers ##
library(tidyverse)
dfDat <- dfGeneralMarkers
dfDat <- unique(dfGeneralMarkers[,c("gene", "cluster")])
if (Obio@parameterList$geneIDcolumn != "hgnc_symbol" & Obio@parameterList$geneIDcolumn != "mgi_symbol"){
refGeneIDcolumn <- "hgnc_symbol"
dfAnno <- Obio@dfGeneAnnotation
dfAnno <- unique(dfAnno[,c("hgnc_symbol",Obio@parameterList$geneIDcolumn )])
dfAnno <- dfAnno[dfAnno[,Obio@parameterList$geneIDcolumn] %in% dfDat[,"gene"],]
dfDat <- merge(
dfDat,
dfAnno,
by.x = "gene",
by.y = Obio@parameterList$geneIDcolumn
)
dfDat$gene <- NULL
names(dfDat) <- gsub("hgnc_symbol", "gene",names(dfDat))
} else {
refGeneIDcolumn <- Obio@parameterList$geneIDcolumn
}
library(dplyr)
dfDat <- dfDat %>%
group_by(cluster) %>%
mutate(rn = row_number()) %>%
ungroup %>%
pivot_wider(names_from = cluster, values_from = gene, values_fill = "") %>% dplyr::select(-rn)
orderVec <- sort(names(dfDat))
dfDat <- dfDat[,orderVec]
names(dfDat) <- paste0(
Obio@parameterList$project_id,
"_Marker_Genes_Cluster_",
names(dfDat)
)
dfDat <- data.frame(dfDat)
## Insert description row ##
descriptionRow <- dfDat[1,]
descriptionRow[1,] <- t(names(dfDat))
descriptionRow[1,] <- paste0(descriptionRow[1,], " from Seurat FindAllMarkers.")
dfDat <- rbind(
descriptionRow,
dfDat
)
# dfCatRef <- names(dfCatRef)
# names(dfCatRef) <- names(dfDat)
#
# dfDat <- rbind(
# dfCatRef,
# dfDat
# )
#######################################################################
## Upload/update category by category ##
updatedCatIDs <- as.vector(NULL, mode = "character")
updatedCatNames <- as.vector(NULL, mode = "character")
for (i in 1:ncol(dfDat)){
cat.name <- names(dfDat)[i]
cat_type <- paste0("temp_cluster_marker_", Obio@parameterList$project_id)
cat.description.text <- as.vector(dfDat[1,i])
gene.vec <- as.vector(
dfDat[,i]
)[2:nrow(dfDat)]
gene.vec <- gene.vec[gene.vec != ""]
gene.vec <- sort(na.omit(gene.vec))
## Determine if cat exists ##
catID <- add.category.to.lab.reference.table.hs(
host = Obio@dbDetailList$host,
pwd = db.pwd,
user = Obio@dbDetailList$db.user,
cat.ref.db = Obio@dbDetailList$ref.cat.db,
cat.ref.db.table = Obio@parameterList$lab.categories.table,
gene.vector = gene.vec,
gene.id = refGeneIDcolumn, #options hgnc_symbol, mgi_symbol
mm.hs.conversion.file = paste0(hpc.mount, "Projects/reference_data/20160303.homologene.data.txt"),
cat_name = cat.name,
cat_type = cat_type,
data_source = paste0(Obio@parameterList$labname, " Lab"),
comments_1 = "",
comments_2 = "",
new.lab.category.table = FALSE,
cat.description.db = "internal_categories",
cat.description.db.table = "category_description",
cat.description.text = cat.description.text,
lab.name = Obio@parameterList$labname,
replaceExistingCatName = TRUE
)
updatedCatIDs <- c(
updatedCatIDs,
catID
)
updatedCatNames <- c(
updatedCatNames,
cat.name
)
} ## End dfDat loop
dfFAMplotIDs <- data.frame(
cat_id = updatedCatIDs,
cat_name = updatedCatNames
)
## Done uploading general markers ##
###############################################################################
if (length(plotList) > 3){
tabVar <- ".tabset .tabset-fade .tabset-dropdown"
} else {
tabVar <- ".tabset .tabset-fade .tabset-pills"
}Dimensionality Reduction Plots by Cluster
cat(paste(knit(text = chnkVec, quiet = T), collapse = '\n'))UMAP_All_Samples
tSNE_All_Samples
UMAP_plot_by_mliGlia
Figure 10: Sample-level UMAPs. Download a pdf of this figure here.
UMAP_plot_by_mliNeuron
Figure 11: Sample-level UMAPs. Download a pdf of this figure here.
Cluster Dendrogram
Figure 12: Clusterplot dendrogram. A pdf of this figure can be downloaded here.
Check Accuracy UMAP Distances
The sleepwalk tool will provide the euclidean distances between individual cells. This will help you to determine clustering accuracy. Find more information here.library(sleepwalk)
sleepwalk(
OsC@reductions$umap@cell.embeddings,
OsC@reductions$pca@cell.embeddings,
saveToFile=paste(Obio@parameterList$outputDir,"sleepwalk.UMAP.html",sep='')
)
htmltools::includeHTML(paste(Obio@parameterList$outputDir,"sleepwalk.UMAP.html",sep=''))## Add UMAP coordinates to Metadata ##
dfAdd <- data.frame(OsC@reductions$umap@cell.embeddings)
OsC <- addDf2seuratMetaData(
obj = OsC,
dfAdd = dfAdd
)
## Add tSNE coordinates to Metadata ##
dfAdd <- data.frame(OsC@reductions$tsne@cell.embeddings)
OsC <- addDf2seuratMetaData(
obj = OsC,
dfAdd = dfAdd
)
plotList <- list()
chnkVec <- as.vector(NULL, mode = "character")
dfTemp <- OsC@meta.data
pos <- grep("Doublet_plot", names(dfTemp))
if (length(pos) > 0){
## First make variation plot for integrated samples, than for all individual samples separately
tag <- "Doublet_plot"
dfTemp$DF_Classification <- factor(dfTemp$DF_Classification, levels = sort(unique(dfTemp$DF_Classification)))
#dotsize <- round(7500/nrow(dfTemp),1)
# dotsize <- 0.3
plotList[[tag]] <- ggplot(dfTemp, aes(UMAP_1, UMAP_2, color=DF_Classification)
)+ geom_point(
shape = 16,
size = as.numeric(dotsize)
) + xlab("UMAP1") + ylab("UMAP2"
) + theme_bw(
) + theme(
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12),
panel.background = element_rect(fill = "lightgrey")
) + ggtitle("Potential Doublets"
) + scale_color_manual(values=c("#000000","#FF0000")
)
#+ xlim(minX, maxX) + ylim(minY, maxY)
## Save to file ##
FNbase <- paste0("DoubletFinderAll", VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
## Create R markdown chunk ##
figLegend <- paste0(
'**Figure ',
figureCount,
'**: Figure depicting the location of potential doublets in PCA components 1 and 2. Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>. '
)
figureCount <- figureCount + 1
NewChnk <- paste0(
" #### Doublets All Timepoints",
"\n```{r ", tag, ", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
}
if (length(plotList) > 3){
tabVar <- ".tabset .tabset-fade .tabset-dropdown"
} else {
tabVar <- ".tabset .tabset-fade .tabset-pills"
}Highlight Potential Doublets
This analyis gives an indication of which cells MIGHT be doublets. Chris McGinnis DoubletFinder package is used to identify potential doublets.
## plot list will be integrated in full figure ##
if (length(pos) > 0){
cat(paste(knit(text = chnkVec, quiet = T), collapse = '\n'))
}###############################################################################
## Estimate cell cycle genes ##
exprFN <- paste0(hpc.mount, "Projects/reference_data/cell_cycle_vignette_files/nestorawa_forcellcycle_expressionMatrix.txt")
exp.mat <- read.table(file = exprFN, header = TRUE,
as.is = TRUE, row.names = 1)
# A list of cell cycle markers, from Tirosh et al, 2015, is loaded with Seurat. We can
# segregate this list into markers of G2/M phase and markers of S phase
s.genes <- cc.genes$s.genes
g2m.genes <- cc.genes$g2m.genes
print(paste0("Used as S-phase marker genes: ", sort(unique(paste(s.genes, collapse = ", ")))))
print(paste0("Used as G2M-phase marker genes: ", sort(unique(paste(g2m.genes, collapse = ", ")))))
# Create our Seurat object and complete the initalization steps
OsC <- CellCycleScoring(OsC, s.features = s.genes, g2m.features = g2m.genes, set.ident = TRUE)
reductionVec <- c("umap", "tsne")
plotList <- list()
chnkVec <- as.vector(NULL, mode = "character")
###############################################################################
## First UMAP all samples together ##
tag <- paste0("CellCyclePhase_All_Samples")
dfPlot <- OsC@meta.data
pos <- grep("included", names(dfPlot))
if (length(pos) == 0){
dfPlot[["included"]] <- "+"
}
dfPlot[["cellID"]] <- row.names(dfPlot)
dfPlot$UMAP_1 <- NULL
dfPlot$UMAP_2 <- NULL
## Get UMAP coordinates ##
coord <- data.frame(OsC@reductions$umap@cell.embeddings)
coord[["cellID"]] <- row.names(coord)
coord <-coord[coord$cellID %in% dfPlot$cellID, ]
dfPlot <- merge(dfPlot, coord, by.x = "cellID", by.y="cellID", all=T)
dfPlot[is.na(dfPlot)] <- 0
dfPlot <- dfPlot[dfPlot$UMAP_1 != 0 & dfPlot$UMAP_2 != 0,]
## Add cluster colors ##
#dfPlot[["Cluster"]] <- paste0("C", dfPlot$seurat_clusters)
#clusterVec <- as.vector(paste0("C", unique(sort(dfPlot$seurat_clusters))))
maxX <- 1.1*max(dfPlot$UMAP_1, na.rm = T)
minX <- 1.1*min(dfPlot$UMAP_1, na.rm = T)
maxY <- 1.1*max(dfPlot$UMAP_2, na.rm = T)
minY <- 1.1*min(dfPlot$UMAP_2, na.rm = T)
# library(scales)
# clusterCols = hue_pal()(length(clusterVec))
# dfPlot$Cluster <- factor(dfPlot$Cluster, levels = clusterVec)
# dotsize = 1
# if (nrow(dfPlot) > 10000){
# dotsize = 0.75
# } else if (nrow(dfPlot) > 20000){
# dotsize = 0.5
# } else if (nrow(dfPlot) > 50000){
# dotsize = 0.25
# }
plotList[[tag]] <- ggplot(data=dfPlot[dfPlot$included == "+",], aes(UMAP_1, UMAP_2, color=Phase)
) + geom_point( shape=16, size = as.numeric(dotsize)
) + xlab("UMAP1") + ylab("UMAP2"
) + theme_bw(
) + theme(
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12),
legend.title = element_blank()
) + ggtitle(paste0("Sample: ", tag)
) + xlim(minX, maxX) + ylim(minY, maxY
) + coord_fixed(ratio=1
)
if (length(unique(dfPlot$Cluster)) > 15){
plotList[[tag]] <- plotList[[tag]] + theme(legend.position = "none")
}
FNbase <- paste0(tag, VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
figLegend <- paste0(
'**Figure ',
figureCount,
':** ',
' UMAP showing all cells from all samples together with the estimated cell-cycle phase color-coded. Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>.'
)
figureCount <- figureCount + 1
NewChnk <- paste0(
"#### ", tag,
"\n```{r CC_UMAP_",
tag,", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
## Done first umap all samples ##
###############################################################################
###############################################################################
## First tsne all samples together ##
tag <- paste0("tSNE_All_Samples")
dfPlot <- OsC@meta.data
pos <- grep("included", names(dfPlot))
if (length(pos) == 0){
dfPlot[["included"]] <- "+"
}
dfPlot[["cellID"]] <- row.names(dfPlot)
dfPlot$tSNE_1 <- NULL
dfPlot$tSNE_2 <- NULL
## Get tsNE coordinates ##
coord <- data.frame(OsC@reductions$tsne@cell.embeddings)
coord[["cellID"]] <- row.names(coord)
coord <-coord[coord$cellID %in% dfPlot$cellID, ]
dfPlot <- merge(dfPlot, coord, by.x = "cellID", by.y="cellID", all=T)
dfPlot[is.na(dfPlot)] <- 0
dfPlot <- dfPlot[dfPlot$tSNE_1 != 0 & dfPlot$tSNE_2 != 0,]
## Add cluster colors ##
#dfPlot[["Cluster"]] <- paste0("C", dfPlot$seurat_clusters)
#clusterVec <- as.vector(paste0("C", unique(sort(dfPlot$seurat_clusters))))
maxX <- 1.1*max(dfPlot$tSNE_1, na.rm = T)
minX <- 1.1*min(dfPlot$tSNE_1, na.rm = T)
maxY <- 1.1*max(dfPlot$tSNE_2, na.rm = T)
minY <- 1.1*min(dfPlot$tSNE_2, na.rm = T)
#library(scales)
#clusterCols = hue_pal()(length(clusterVec))
#dfPlot$Cluster <- factor(dfPlot$Cluster, levels = clusterVec)
# dotsize = 1.5
# if (nrow(dfPlot) > 10000){
# dotsize = 0.75
# } else if (nrow(dfPlot) > 50000){
# dotsize = 0.5
# } else {
# dotsize = 0.25
# }
plotList[[tag]] <- ggplot(data=dfPlot[dfPlot$included == "+",], aes(tSNE_1, tSNE_2, color=Phase)
) + geom_point( shape=16, size = as.numeric(dotsize)
) + xlab("tSNE1") + ylab("tSNE2"
) + theme_bw(
) + theme(
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12),
legend.title = element_blank()
) + ggtitle(paste0("Sample: ", tag)
) + xlim(minX, maxX) + ylim(minY, maxY
) + coord_fixed(ratio=1
)
if (length(unique(dfPlot$Cluster)) > 15){
plotList[[tag]] <- plotList[[tag]] + theme(legend.position = "none")
}
FNbase <- paste0(tag, VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
figLegend <- paste0(
'**Figure ',
figureCount,
':** ',
' tSNE showing all cells from all samples together. The esimated cell-cycle phase is color coded. Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>.'
)
figureCount <- figureCount + 1
NewChnk <- paste0(
"#### ", tag,
"\n```{r CC_tSNE_",
tag,", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
## Done first tsne all samples ##
###############################################################################
###############################################################################
## Make one UMAP plot per sample ##
sampleVec <- sort(unique(OsC@meta.data$sampleID))
dfPlot <- OsC@meta.data
pos <- grep("included", names(dfPlot))
if (length(pos) == 0){
dfPlot[["included"]] <- "+"
}
dfPlot[["cellID"]] <- row.names(dfPlot)
## Get UMAP coordinates ##
coord <- data.frame(OsC@reductions$umap@cell.embeddings)
coord[["cellID"]] <- row.names(coord)
coord <-coord[coord$cellID %in% dfPlot$cellID, ]
dfPlot$UMAP_1 <- NULL
dfPlot$UMAP_2 <- NULL
dfPlot <- merge(dfPlot, coord, by.x = "cellID", by.y="cellID", all=T)
dfPlot[is.na(dfPlot)] <- 0
dfPlot <- dfPlot[dfPlot$UMAP_1 != 0 & dfPlot$UMAP_2 != 0,]
## Add cluster colors ##
#dfPlot[["Cluster"]] <- paste0("C", dfPlot$seurat_clusters)
#clusterVec <- as.vector(paste0("C", unique(sort(dfPlot$seurat_clusters))))
#library(scales)
#clusterCols = hue_pal()(length(clusterVec))
#dfPlot$Cluster <- factor(dfPlot$Cluster, levels = clusterVec)
maxX <- 1.1*max(dfPlot$UMAP_1, na.rm = T)
minX <- 1.1*min(dfPlot$UMAP_1, na.rm = T)
maxY <- 1.1*max(dfPlot$UMAP_2, na.rm = T)
minY <- 1.1*min(dfPlot$UMAP_2, na.rm = T)
for (i in 1:length(sampleVec)){
tag <- paste0("UMAP_CC_plot_by_", sampleVec[i])
dfPlotSel <- dfPlot[dfPlot$sampleID == sampleVec[i], ]
plotList[[tag]] <- ggplot(data=dfPlotSel[dfPlot$included == "+",], aes(UMAP_1, UMAP_2, color=Phase)
) + geom_point( shape=16, size = as.numeric(dotsize)
) + xlab("UMAP1") + ylab("UMAP2"
) + theme_bw(
) + theme(
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12),
legend.title = element_blank()
) + ggtitle(paste0("Sample: ", tag)
) + xlim(minX, maxX) + ylim(minY, maxY
) + coord_fixed(ratio=1
)
if (length(unique(dfPlot$Cluster)) > 15){
plotList[[tag]] <- plotList[[tag]] + theme(legend.position = "none")
}
FNbase <- paste0(tag, VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
figLegend <- paste0(
'**Figure ',
figureCount,
':** ',
' Sample-level UMAPs. Estimated cell-cylce phase color-coded. Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>.'
)
figureCount <- figureCount + 1
NewChnk <- paste0(
paste("#### ", tag),
"\n```{r CC_UMAP_",
tag,", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
}
## Done making one umap plot per sample ##
###############################################################################
###############################################################################
## Add cluster dendrogram by sample ##
if (length(unique(OsC@meta.data$sampleID)) > 3){
library(ggtree)
Idents(OsC) <- "sampleID"
OsC <- BuildClusterTree(OsC)
tag <- paste0("Sample_Dendrogram")
OsC@tools$BuildClusterTree$tip.label <- paste0("C", OsC@tools$BuildClusterTree$tip.label)
plotList[[tag]] <- ggplot(OsC@tools$BuildClusterTree
) + geom_tree(
) + theme_tree(
) + geom_tiplab(
) + labs(title=tag
) + theme(
panel.border = element_rect(colour = "black", fill=NA, size=1),
axis.title.x=element_blank(),
plot.title = element_text(hjust = 0.5, size = 12)
) + xlim(0,1.2*max(OsC@tools$BuildClusterTree[[2]]))
## Save to file ##
FNbase <- paste0(tag,".", VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
figLegend <- paste0(
'**Figure ',
figureCount,
':** ',
' Clusterplot dendrogram by sample ID. ','A pdf of this figure can be downloaded <a href="',FNrel,'", target="_blank>here</a>.'
)
NewChnk <- paste0(
"#### SampleID Dendrogram",
"\n```{r ", tag, "results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
figureCount <- figureCount + 1
}
## Done by sample ##
###############################################################################
###############################################################################
## Add cluster dendrogram by cluster ##
library(ggtree)
Idents(OsC) <- "seurat_clusters"
OsC <- BuildClusterTree(OsC)
tag <- paste0("Cluster_Dendrogram")
OsC@tools$BuildClusterTree$tip.label <- paste0("C", OsC@tools$BuildClusterTree$tip.label)
plotList[[tag]] <- ggplot(OsC@tools$BuildClusterTree
) + geom_tree(
) + theme_tree(
) + geom_tiplab(
) + labs(title=tag
) + theme(
panel.border = element_rect(colour = "black", fill=NA, size=1),
axis.title.x=element_blank(),
plot.title = element_text(hjust = 0.5, size = 12)
) + xlim(0,1.2*max(OsC@tools$BuildClusterTree[[2]]))
#+ xlim(-1,1.2*max(OsC@tools$BuildClusterTree$edge))
## Save to file ##
FNbase <- paste0(tag,".", VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
figLegend <- paste0(
'**Figure ',
figureCount,
':** ',
' Clusterplot dendrogram. ','A pdf of this figure can be downloaded <a href="',FNrel,'" target="_blank">here</a>.'
)
NewChnk <- paste0(
"#### Cluster Dendrogram",
"\n```{r ", tag, "results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
figureCount <- figureCount + 1
## Done integraed analysis ##
###############################################################################
if (length(plotList) > 3){
tabVar <- ".tabset .tabset-fade .tabset-dropdown"
} else {
tabVar <- ".tabset .tabset-fade .tabset-pills"
}Cell Cycle Markers
cat(paste(knit(text = chnkVec, quiet = T), collapse = '\n'))CellCyclePhase_All_Samples
Figure 13: UMAP showing all cells from all samples together with the estimated cell-cycle phase color-coded. Download a pdf of this figure here.
tSNE_All_Samples
Figure 14: tSNE showing all cells from all samples together. The esimated cell-cycle phase is color coded. Download a pdf of this figure here.
UMAP_CC_plot_by_mliGlia
Figure 15: Sample-level UMAPs. Estimated cell-cylce phase color-coded. Download a pdf of this figure here.
UMAP_CC_plot_by_mliNeuron
Figure 16: Sample-level UMAPs. Estimated cell-cylce phase color-coded. Download a pdf of this figure here.
Cluster Dendrogram
Figure 17: Clusterplot dendrogram. A pdf of this figure can be downloaded here.
###############################################################################
## Create datatable for plotting ##
## This plotting procedure requires three sets: the sets cellTypeIDs, clusterIDs, cellTypeIDs
## level1ID, level2ID, level3ID
sampleIDs <- unique(OsC@meta.data$sampleID)
#Obio@parameterList$singleCellSeuratMtCutoff <- 20
if (is.null(Obio@parameterList$singleCellSeuratMtCutoff)){
Obio@parameterList$singleCellSeuratMtCutoff <- rep(10, length(sampleIDs))
} else if (length(Obio@parameterList$singleCellSeuratMtCutoff) == 1){
Obio@parameterList$singleCellSeuratMtCutoff <- rep(
Obio@parameterList$singleCellSeuratMtCutoff,
length(sampleIDs)
)
} else if (length(Obio@parameterList$singleCellSeuratMtCutoff) == length(sampleIDs)){
Obio@parameterList$singleCellSeuratMtCutoff <- Obio@parameterList$singleCellSeuratMtCutoff
} else {
print("Can't determine mitochondrial cut off. ")
stop()
}
sampleIDs <- unique(OsC@meta.data$sampleID)
clusterIDs <- unique(OsC@meta.data[,Obio@parameterList$singleCellClusterString])
if (length(grep("cellIdent", names(OsC@meta.data))) == 0){
OsC@meta.data[["cellIdent"]] <- "All"
}
cellTypeIDs <- unique(OsC@meta.data[,"cellIdent"])
dfTemp <- OsC@meta.data
if (length(grep("^cellIdent$", names(dfTemp))) == 0){
dfTemp[["cellIdent"]] <- "All"
}
dfTemp <- dfTemp[dfTemp$percent.mt <= max(Obio@parameterList$singleCellSeuratMtCutoff), ]
dfTemp[["cellID"]] <- row.names(dfTemp)
dfTemp <- unique(dfTemp[,c("cellID", "sampleID", Obio@parameterList$singleCellClusterString, "cellIdent")])
names(dfTemp) <- gsub(Obio@parameterList$singleCellClusterString, "Cluster", names(dfTemp) )
names(dfTemp) <- gsub(Obio@parameterList$singleCellClusterString, "Cluster", names(dfTemp) )
dfTemp <- unique(dfTemp[,c("cellID", "sampleID", "Cluster", "cellIdent")])
dfRes <- dfTemp
dfRes$cellID <- NULL
row.names(dfRes) <- NULL
dfRes <- unique(dfRes)
dfRes[["N_cells"]] <- 0
for (i in 1:nrow(dfRes)){
dfRes[i, "N_cells"] <- nrow(dfTemp[dfTemp$sampleID == dfRes[i,"sampleID"] & dfTemp$Cluster == dfRes[i,"Cluster"] & dfTemp$cellIdent == dfRes[i,"cellIdent"], ])
}
## Calculate cluster percentages per celltypeID ##
dfRes[["Perc_cells"]] <- 0
for (i in 1:length(cellTypeIDs)){
dfResTemp2 <- dfRes[dfRes$cellIdent == cellTypeIDs[i], ]
tempCluster <- as.vector(unique(dfResTemp2$Cluster))
for (j in 1:length(tempCluster)){
dfResTemp3 <- dfResTemp2[dfResTemp2$Cluster == tempCluster[j],]
NclusterTotal <- sum(dfResTemp3[, "N_cells"])
dfResTemp3[,"Perc_cells"] <- round(dfResTemp3[,"N_cells"]/NclusterTotal, 4)*100
if (j ==1){
dfRes3 <- dfResTemp3
} else {
dfRes3 <- rbind(dfResTemp3, dfRes3)
}
}
if (i ==1){
dfRes4 <- dfRes3
} else {
dfRes4 <- rbind(dfRes3, dfRes4)
}
}
dfRes <- dfRes4
plotList <- list()
chnkVec <- as.vector(NULL, mode = "character")
for (i in 1:length(cellTypeIDs)){
#############################################################################
## Create cell number plot ##
tag <- paste0(cellTypeIDs[i], "_Number")
dfResTemp <- dfRes[dfRes$cellIdent == cellTypeIDs[i], ]
## Calculate percentages for this subset ##
plotList[[tag]] <- ggplot(
) + geom_bar(aes(y = N_cells, x = Cluster, fill = sampleID), data = dfResTemp, stat="identity"
) + labs(title=tag, x="Cluster", y = "Cell Count"
) + theme_bw(
) + theme(
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12)
) + coord_flip()
###########################################################################
## Save plot to file ##
FNbase <- paste0(tag,".Ncells", VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
## ##
###########################################################################
###########################################################################
## Add to chunk ##
figCap <- paste0(
'**Figure ',
figureCount,
'A:** Cell Count in each cluster for ',
tag,
'Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>. '
)
NewChnk <- paste0(
paste0("#### Barchart_ ", tag),
"\n```{r Barchart-",tag,", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",figCap,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
## Done adding ##
#############################################################################
#############################################################################
## Add percentage plot ##
tag <- paste0(cellTypeIDs[i], "_Percent")
plotList[[tag]] <- ggplot(
) + geom_bar(aes(x = Cluster, y = Perc_cells, fill = sampleID), data = dfResTemp, stat="identity"
) + labs(title=tag, x="Cluster", y = "Percent Cells"
) + theme_bw(
) + theme(
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12)
) + coord_flip()
###########################################################################
## Save plot to file ##
FNbase <- paste0(tag, ".percent.cells",VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
## ##
###########################################################################
###########################################################################
## Add to chunk ##
figCap <- paste0(
'**Figure ',
figureCount,
'B:** Cell percentages in each cluster for ',
tag,
'Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>. '
)
NewChnk <- paste0(
paste0("#### Barchart_ ", tag),
"\n```{r Barchart-percent_",tag,", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",figCap,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
## Done adding percentage plot ##
#############################################################################
figureCount <- figureCount + 1
}
## Done creating data table ##
###############################################################################
if (length(plotList) > 3){
tabVar <- ".tabset .tabset-fade .tabset-dropdown"
} else {
tabVar <- ".tabset .tabset-fade .tabset-pills"
}Barchart N cells and Percent in Clusters
if you could prepare a bar-shape graph with the %of cells clusters representing our populations (like in the Nat Med)
cat(paste(knit(text = chnkVec, quiet = T), collapse = '\n'))Barchart_ All_Number
Figure 18A: Cell Count in each cluster for All_NumberDownload a pdf of this figure here.
Barchart_ All_Percent
Figure 18B: Cell percentages in each cluster for All_PercentDownload a pdf of this figure here.
# save(Obio,
# file = paste0(
# Obio@parameterList$localWorkDir,
# Obio@parameterList$project_id,
# ".temp.bioLOGIC.Robj"
# )
# )
#print("Obio Object saved.")
save(OsC,
file = paste0(
Obio@parameterList$localWorkDir,
Obio@parameterList$project_id,
".Seurat.Robj"
)
)
library(AUCell)
plotList <- list()
chnkVec <- as.vector(NULL, mode = "character")
# Defined in the section above #
## This needs to become a gmt file ##
if (is.null(Obio@parameterList$catRefFile)){
FNcat <- paste0(hpc.mount, "Projects/schaefera/tobias.ackels/360_scRNAseq_mm_10X_1M_neurons_20k/basedata/asl320.referenceCats.txt")
} else {
FNcat <- Obio@parameterList$catRefFile
}
if (length(grep(".gmt$", FNcat)) > 0){
print("Load gmt file. To be implemented.")
stop()
} else {
dfHeatmapGenes <- read.delim(
FNcat,
header = T,
sep = "\t",
stringsAsFactors = F
)
if (is.null(Obio@parameterList[["cat2DotplotList"]])){
Obio@parameterList[["cat2DotplotList"]] <- list()
}
if (is.null(Obio@parameterList[["cat2HMplotList"]])){
Obio@parameterList[["cat2HMplotList"]] <- list()
}
for (i in 1:ncol(dfHeatmapGenes)){
genes <- as.vector(dfHeatmapGenes[2:nrow(dfHeatmapGenes),i])
genes <- genes[genes %in% rownames(x = OsC@assays$RNA)]
if (length(unique(genes)) < 61 | (length(unique(genes)) > 0)){
Obio@parameterList[["cat2DotplotList"]][[names(dfHeatmapGenes)[i]]] <- genes
}
if ((length(unique(genes)) < 501) | (length(unique(genes)) > 2) ){
Obio@parameterList[["cat2HMplotList"]] [[names(dfHeatmapGenes)[i]]] <- genes
}
}
}
## Add transcription factors to dotplot ##
if (Obio@parameterList$geneIDcolumn != "mgi_symbol" & Obio@parameterList$geneIDcolumn != "hgnc_symbol") {
queryGS <- "hgnc_symbol"
} else {
queryGS <- Obio@parameterList$geneIDcolumn
}
tempVec <- retrieve.gene.category.from.db(
cat_id = "ag_lab_categories__10",
password = db.pwd,
gene.symbol = queryGS,
user = Obio@parameterList$db.user,
host = Obio@parameterList$host
)
###############################################################################
## If this is fish, translation is non-human or non-mouse, translation is necessary
if (queryGS != Obio@parameterList$geneIDcolumn){
dfAnno <- Obio@dfGeneAnnotation
dfAnno <- unique(dfAnno[,c("hgnc_symbol",Obio@parameterList$geneIDcolumn )])
dfAnno <- dfAnno[dfAnno$hgnc_symbol != "", ]
dfAnno <- dfAnno[dfAnno$hgnc_symbol %in% tempVec, ]
tempVec <- unique(dfAnno[,Obio@parameterList$geneIDcolumn])
tempVec <- tempVec[tempVec != ""]
}
dfHMG <- dfGeneralMarkers[dfGeneralMarkers$gene %in% tempVec, ]
dfHMGsel <- data.frame(dfHMG %>% group_by(cluster) %>% top_n(5, avg_diff))
Obio@parameterList[["cat2DotplotList"]][["Top5_TF_per_cluster_Markers"]] <- as.vector(unique(dfHMGsel$gene))
## Add cluster defining transcription factors to the collection ##
## For the dotplot ##
###############################################################################
## Get backdrop
exprMatrix <- as.matrix(OsC@assays$RNA@counts)
#logMat <- log10(exprMatrix+1)
# When using a Seurat object #
logMat <- data.frame(OsC[["RNA"]]@data)
## Load tSNE coordinates ##
cellsTsne <- data.frame(OsC@reductions$umap@cell.embeddings)
## done
FNbase <- paste0("CatScatter_Rankings", VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
cells_rankings <- AUCell_buildRankings(exprMatrix)
dev.off()
geneSets <- Obio@parameterList$cat2DotplotList
cells_AUC <- AUCell_calcAUC(geneSets, cells_rankings, aucMaxRank=nrow(cells_rankings)*0.05)
## Select thresholds ##
FNbase <- paste0("CatScatterHist", VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
set.seed(123)
cells_assignment <- AUCell_exploreThresholds(
cells_AUC,
plotHist=TRUE,
nCores=1,
assign=TRUE
)
dev.off()
## Add data to dfExpr ##
## Plot CatScatters ##
for (i in 1:length(Obio@parameterList$cat2DotplotList)){
HMname <- names(Obio@parameterList$cat2DotplotList)[i]
tag <- gsub("[.]", "_", HMname)
FNbase <- paste0("CatScatterHist_", HMname, VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
selectedThresholds <- cells_assignment[[i]]$aucThr$thresholds
if ("minimumDens" %in% rownames(selectedThresholds)) {
pThr <- selectedThresholds["minimumDens", "threshold"]
} else if ("Global_k1" %in% rownames(selectedThresholds)){
pThr <- selectedThresholds["Global_k1", "threshold"]
} else {
pThr <- selectedThresholds[1, "threshold"]
}
if (nrow(cellsTsne) > 15000){
cex = 0.25
} else if (nrow(cellsTsne) > 1000){
cex = 0.5
} else {
cex = 1
}
## Get AUC matrix ##
tSNE.df <- data.frame(cellsTsne, cell=rownames(cellsTsne))
mAUC <- getAUC(cells_AUC)[HMname,rownames(tSNE.df)]
dfAUC <- data.frame(mAUC)
dfAUC[["cellID"]] <- row.names(dfAUC)
dfAUC <- merge(dfAUC, tSNE.df, by.x = "cellID", by.y = "cell")
dfDocAUC <- unique(dfAUC[,c("cellID", "mAUC")])
dfDocAUC[["cat"]] <- paste0("Cat_",tag)
if (i == 1){
dfResAUC <- dfDocAUC
} else {
dfResAUC <- rbind(
dfResAUC,
dfDocAUC
)
}
input <- list(
"x_axis" = "UMAP1",
"y_axis" = "UMAP2",
"gene" = HMname
)
#dotsize <- cex
legendNote <- paste0(
" The following genes of this dataset are represented in this figure: ",
paste0(sort(Obio@parameterList$cat2DotplotList[[i]]), collapse = ", ")
)
plotList[[tag]] <- ggplot(data = dfAUC, aes(x=UMAP_1, y=UMAP_2, color = mAUC)
)+ geom_point( shape=16, size = dotsize
) + scale_color_gradient("AUC", low="grey", high="darkblue"
) + xlab(input$x_axis) + ylab(input$y_axis
) + theme_bw(
) + theme(
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12)
)+ ggtitle(paste0("Category: ", input$gene)
) + coord_fixed(ratio = 1
)
#+ theme(legend.position="none")
FNbase <- paste0("CatScatter", HMname, VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
## Create R markdown chunk ##
figLegend <- paste0(
'**Figure ',
figureCount,
'A:** Category Scatter showing gene category ',
HMname, '. ', legendNote,
'. Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>. '
)
NewChnk <- paste0(
"#### Category Feature Plot ",HMname,
"\n```{r CatFeatPlot1_",
i,", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
###########################################################################
## Add part B - dotplot ##
DefaultAssay(OsC) <- "RNA"
OsC@meta.data[["hmIdent2"]] <- paste0("C", OsC@meta.data[,Obio@parameterList$singleCellClusterString])
levels <- paste0(
"C",
sort(unique(OsC@meta.data[,Obio@parameterList$singleCellClusterString]))
)
OsC@meta.data$hmIdent2 <- factor(OsC@meta.data$hmIdent2, levels=levels)
Idents(OsC) <- "hmIdent2"
HMname <- paste0("Dotplot_", names(Obio@parameterList$cat2DotplotList)[i])
tag <- gsub("[.]", "_", HMname)
dpGenes <- unique(Obio@parameterList$cat2DotplotList[[i]])
legendNote <- paste0("The following genes were found in this category and the single-cell dataset: ", paste0(dpGenes, collapse=", "))
plotList[[tag]] <- DotPlotSB(
object = OsC,
features = dpGenes,
#cols = cols,
group.by = NULL,
split.by = NULL,
dot.scale = 4,
col.min = 0,
col.max = 5,
#assay = "RNA"
) + ggtitle(gsub("_", "", tag)) + coord_fixed() + coord_flip()
FNbase <- paste0(HMname, VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
## Create R markdown chunk ##
figLegend <- paste0(
'**Figure ',
figureCount,
'B:** Dotplot showing gene category ',
HMname, '. ', legendNote,
'. Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>. '
)
figureCount <- figureCount + 1
NewChnk <- paste0(
"\n```{r ",tag,
", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
## Done adding dotplot ##
###########################################################################
}
Obio@dataTableList[["dfResAUC"]] <- dfResAUC
if (length(plotList) > 3){
tabVar <- ".tabset .tabset-fade .tabset-dropdown"
} else {
tabVar <- ".tabset .tabset-fade .tabset-pills"
}Category Enrichment Scatterplots
cat(paste(knit(text = chnkVec, quiet = T), collapse = '\n'))Category Feature Plot Enteric_Glia
Figure 19A: Category Scatter showing gene category Enteric_Glia. The following genes of this dataset are represented in this figure: S100b, Sox10. Download a pdf of this figure here.
Figure 19B: Dotplot showing gene category Dotplot_Enteric_Glia. The following genes were found in this category and the single-cell dataset: S100b, Sox10. Download a pdf of this figure here.
Category Feature Plot Enteric_Neuron
Figure 20A: Category Scatter showing gene category Enteric_Neuron. The following genes of this dataset are represented in this figure: Chat, Elavl3, Elavl4, Gad1, Nos1, Ret. Download a pdf of this figure here.
Figure 20B: Dotplot showing gene category Dotplot_Enteric_Neuron. The following genes were found in this category and the single-cell dataset: Ret, Chat, Elavl3, Elavl4, Gad1, Nos1. Download a pdf of this figure here.
Category Feature Plot Proliferation
Figure 21A: Category Scatter showing gene category Proliferation. The following genes of this dataset are represented in this figure: Mki67, Top2a. Download a pdf of this figure here.
Figure 21B: Dotplot showing gene category Dotplot_Proliferation. The following genes were found in this category and the single-cell dataset: Mki67, Top2a. Download a pdf of this figure here.
Category Feature Plot Apoptosis
Figure 22A: Category Scatter showing gene category Apoptosis. The following genes of this dataset are represented in this figure: Acin1, Acvr1c, Aifm3, Akt1, Apc, Bbc3, Bcap31, Birc2, Blcap, Bmx, Bnip1, Capn10, Casp2, Casp3, Casp6, Casp7, Casp8, Cdk5rap3, Cecr2, Cflar, Cideb, Cidec, Clspn, Dedd2, Dffa, Dffb, Dicer1, Dnase1l3, Dnase2a, Dnm1l, Endog, Ern2, Fnta, H1f0, Hmgb2, Htra2, Kpna1, Kpnb1, Madd, Prkcd, Prkcq, Ptk2, Rock1, Satb1, Sharpin, Stk24, Taok1, Tardbp, Top2a, Xkr8. Download a pdf of this figure here.
Figure 22B: Dotplot showing gene category Dotplot_Apoptosis. The following genes were found in this category and the single-cell dataset: Acin1, Acvr1c, Aifm3, Akt1, Apc, Bbc3, Bcap31, Birc2, Blcap, Bmx, Bnip1, Capn10, Casp2, Casp3, Casp6, Casp7, Casp8, Cdk5rap3, Cecr2, Cflar, Cideb, Cidec, Clspn, Dedd2, Dffa, Dffb, Dicer1, Dnase1l3, Dnase2a, Dnm1l, Endog, Ern2, Fnta, H1f0, Hmgb2, Htra2, Kpna1, Kpnb1, Madd, Prkcd, Prkcq, Ptk2, Rock1, Satb1, Sharpin, Stk24, Taok1, Tardbp, Top2a, Xkr8. Download a pdf of this figure here.
Category Feature Plot Top5_TF_per_cluster_Markers
Figure 23A: Category Scatter showing gene category Top5_TF_per_cluster_Markers. The following genes of this dataset are represented in this figure: Ahr, Casz1, Creb5, Cux2, Dach1, Ebf1, Ebf3, Etv1, Hoxb5, Lrrfip1, Myt1l, Nfatc1, Nfia, Npas3, Pbx3, Phox2b, Rarb, Rfx3, Rora, Satb1, Satb2, Sox5, Sox6, St18, Tbx3, Thrb, Tlx2, Trps1, Tshz2, Tshz3, Zbtb20, Zfhx4, Zfp521, Zfp804a. Download a pdf of this figure here.
Figure 23B: Dotplot showing gene category Dotplot_Top5_TF_per_cluster_Markers. The following genes were found in this category and the single-cell dataset: Myt1l, Lrrfip1, Zbtb20, Zfp804a, Rora, Sox5, Nfia, Npas3, Sox6, Cux2, Phox2b, Tlx2, Ahr, Rarb, Etv1, Tbx3, Ebf1, Casz1, Zfp521, Tshz2, Satb1, Pbx3, Trps1, Dach1, Ebf3, Rfx3, Nfatc1, Tshz3, Hoxb5, St18, Thrb, Satb2, Creb5, Zfhx4. Download a pdf of this figure here.
## Get average Expression for all cells
dfAvgExprAllCells <- Obio@dataTableList$dfAvglg10ExprAll
## Get average expression per cluster
dfAvgExprByCluster <- Obio@dataTableList$dfAvglg10ExprPerCluster
## Merge
dfDataTable <- merge(
dfAvgExprAllCells,
dfAvgExprByCluster,
by.x = "gene",
by.y = "gene",
all =T
)
dfDataTable[is.na(dfDataTable)] <- 0
clusterVec <- names(dfDataTable)
clusterVec <- clusterVec[!(clusterVec %in% c("gene", "all"))]
slopeVec <- as.vector(NULL, mode = "numeric")
intersectVec <- as.vector(NULL, mode = "numeric")
for (i in 1:length(clusterVec)){
LMformula <- as.formula(paste0(clusterVec[i]," ~ all"))
clusterFit <- lm(data=dfDataTable, formula = LMformula)
slopeVec <- c(
slopeVec,
coef(clusterFit)[[2]]
)
intersectVec <- c(
intersectVec,
coef(clusterFit)[[1]]
)
residuals <- round(clusterFit$residuals, 3)
dfTempResiduals <- data.frame(dfDataTable$gene, residuals)
names(dfTempResiduals) <- c("gene", paste0(clusterVec[i], "_Residuals"))
if (i ==1){
dfClusterResiduals <- dfTempResiduals
} else {
dfClusterResiduals <- merge(
dfClusterResiduals,
dfTempResiduals,
by.x = "gene",
by.y = "gene",
all =T
)
}
dfClusterResiduals[is.na(dfClusterResiduals)] <- 0
}
Obio@dataTableList[["dfClusterResiduals"]] <- dfClusterResiduals
## Make gene set with Residual marker genes ##
if (Obio@parameterList$geneIDcolumn != "hgnc_symbol" & Obio@parameterList$geneIDcolumn != "mgi_symbol"){
refGeneIDcolumn <- "hgnc_symbol"
dfAnno <- Obio@dfGeneAnnotation
dfAnno <- unique(dfAnno[,c("hgnc_symbol",Obio@parameterList$geneIDcolumn )])
dfAnno <- dfAnno[dfAnno[,Obio@parameterList$geneIDcolumn] %in% dfClusterResiduals[,"gene"],]
dfClusterResiduals <- merge(
dfClusterResiduals,
dfAnno,
by.x = "gene",
by.y = Obio@parameterList$geneIDcolumn
)
dfClusterResiduals$gene <- NULL
names(dfClusterResiduals) <- gsub("hgnc_symbol", "gene",names(dfClusterResiduals))
} else {
refGeneIDcolumn <- Obio@parameterList$geneIDcolumn
}
residualClusterMarkers <- list()
DefaultcutOff <- 1
clusterVec <- names(dfClusterResiduals)
clusterVec <- clusterVec[clusterVec != "gene"]
for (i in 1:length(clusterVec)){
clusterGenes <- as.vector(sort(unique(dfClusterResiduals[dfClusterResiduals[, clusterVec[i]] > DefaultcutOff,"gene"])))
if (length(clusterGenes) < 3){
usedCutOff <- DefaultcutOff/2
clusterGenes <- as.vector(sort(unique(dfClusterResiduals[dfClusterResiduals[, clusterVec[i]] > usedCutOff,"gene"])))
} else {
usedCutOff <- DefaultcutOff
}
if (length(clusterGenes) > 1){
residualClusterMarkers[[paste0(clusterVec[[i]], "_Linear_Regression_Markers_", DefaultcutOff)]] <- c(
paste0(clusterVec[[i]], "_Linear_Regression_Markers_", gsub("\\.", "_", usedCutOff) , " Experiment ", Obio@parameterList$project_id),
clusterGenes
)
}
}
## Upload Marker Genes for this Project ##
#######################################################################
## Upload/update category by category ##
updatedCatIDs <- as.vector(NULL, mode = "character")
updatedCatNames <- as.vector(NULL, mode = "character")
for (i in 1:length(residualClusterMarkers)){
cat.name <- names(residualClusterMarkers)[i]
cat_type <- paste0("temp_cluster_marker_", Obio@parameterList$project_id)
cat.description.text <- as.vector(residualClusterMarkers[[i]][1])
gene.vec <- as.vector(
residualClusterMarkers[[i]]
)[2:length(residualClusterMarkers[[i]])]
gene.vec <- gene.vec[gene.vec != ""]
gene.vec <- sort(na.omit(gene.vec))
## Determine if cat exists ##
catID <- add.category.to.lab.reference.table.hs(
host = Obio@dbDetailList$host,
pwd = db.pwd,
user = Obio@dbDetailList$db.user,
cat.ref.db = Obio@dbDetailList$ref.cat.db,
cat.ref.db.table = Obio@parameterList$lab.categories.table,
gene.vector = gene.vec,
gene.id = refGeneIDcolumn, #options hgnc_symbol, mgi_symbol
mm.hs.conversion.file = paste0(hpc.mount, "Projects/reference_data/20160303.homologene.data.txt"),
cat_name = cat.name,
cat_type = cat_type,
data_source = paste0(Obio@parameterList$labname, " Lab"),
comments_1 = "",
comments_2 = "",
new.lab.category.table = FALSE,
cat.description.db = "internal_categories",
cat.description.db.table = "category_description",
cat.description.text = cat.description.text,
lab.name = Obio@parameterList$labname,
replaceExistingCatName = TRUE
)
updatedCatIDs <- c(
updatedCatIDs,
catID
)
updatedCatNames <- c(
updatedCatNames,
cat.name
)
} ## End dfDat loop
dfCatNameLR <- data.frame(catID = updatedCatIDs, cat_name= updatedCatNames)
Obio@dataTableList[["residualClusterMarkers"]] <- residualClusterMarkers
###############################################################################
## Creaate Plot array ##
plotList <- list()
chnkVec <- as.vector(NULL, mode = "character")
# x-axis: all
# y-axis: intensity cluster X
# highlight: Marker genes
dfClusterResiduals <- Obio@dataTableList$dfClusterResiduals
clusterVec <- names(dfDataTable)
clusterVec <- clusterVec[!(clusterVec %in% c("gene", "all"))]
## Determine cluster colors ##
dfClustCol <- data.frame(cluster=clusterVec, clusterCol=clusterVec)
dfClustCol$clusterCol <- as.numeric(gsub("C_", "", dfClustCol$clusterCol))
dfClustCol <- dfClustCol[order(dfClustCol$clusterCol, decreasing = F),]
library(scales)
clusterCols = hue_pal()(nrow(dfClustCol))
dfClustCol[["clusterCol"]] <- clusterCols
clusterVec <- as.vector(dfClustCol$cluster)
clusterCol <- as.vector(dfClustCol$clusterCol)
for (i in 1:length(clusterVec)){
## Determine top10 markers ##
clustCol <- names(dfClusterResiduals)[grep(paste0(clusterVec[i],"_"), names(dfClusterResiduals))]
dfClusterResiduals <- dfClusterResiduals[order(dfClusterResiduals[,clustCol], decreasing = T), ]
markerVec <- as.vector(dfClusterResiduals[1:10,"gene"])
dfTempResiduals <- dfClusterResiduals[dfClusterResiduals[,clustCol] > DefaultcutOff, ]
highlightGenes <- as.vector(dfTempResiduals[,"gene"])
dfPlot <- dfDataTable[,c("gene", "all", clusterVec[i])]
dfPlot[["label"]] <- ""
dfPlot[dfPlot$gene %in% markerVec, "label"] <- dfPlot[dfPlot$gene %in% markerVec, "gene"]
## Add highlight ##
dfPlot[["ClusterMarker"]] <- ""
dfPlot[dfPlot$gene %in% highlightGenes, "ClusterMarker"] <- "+"
dfPlot[["x"]] <- dfPlot$all
dfPlot[["y"]] <- dfPlot[, clusterVec[i]]
tag <- paste0("LinearRegressionMarkers_Cluster", clusterVec[i])
## Create fitted line ##
lineFit <- lm(data=dfPlot, formula = y ~ x)
library(ggrepel)
plotList[[tag]] <- ggplot(
data = dfPlot,
aes(
x=x,
y=y, label = label, color = ClusterMarker)
#) + geom_abline(intercept = intersectVec[i], slope = slopeVec[i], linetype="dashed", color = "grey"
) + geom_smooth(method='lm', formula= y ~ x, linetype="dashed", se = T, colour = "grey"
) + geom_point( shape=16, size = dotsize
) + xlab("Average Expression All Cells") + ylab(paste0("Average Expression ", clusterVec[[i]])
) + theme_bw(
) + theme(
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12)
) + ggtitle(paste0("Cluster Markers ", clusterVec[[i]])
) + scale_color_manual(values=c("#000000", clusterCol[i])
) + geom_text_repel(
)
# + xlim(0, xmax) + ylim(0, ymax)
###########################################################################
## Save plot to file ##
FNbase <- paste0("Linear.regression.cluster.markers.",clusterVec[i], VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
## ##
###########################################################################
pos <- grep(paste0(clusterVec[i],"_"), dfCatNameLR$cat_name)
LRcatID <- ""
if (length(pos) == 1){
LRcatID <- as.vector(dfCatNameLR[pos, "catID"])
}
clusterVec <- as.vector(dfClustCol$cluster)
clusterCol <- as.vector(dfClustCol$clusterCol)
pos <- grep(paste0(gsub("C_", "Cluster_",clusterVec[i]),"$"), dfFAMplotIDs$cat_name)
DGEcatID <- ""
if (length(pos) == 1){
DGEcatID <- as.vector(dfFAMplotIDs[pos, "cat_id"])
}
link <- paste0('<a href="https://',urlString,'/',Obio@parameterList$project_id,'/scatterplot?x_axis=add_counts_Avg_log10_Expr_all&y_axis=add_counts_Avg_log10_Expr_',clusterVec[i], '&cat_id=', LRcatID,'" target="_blank">here</a>')
linkDGE <- paste0('<a href="https://',urlString,'/',Obio@parameterList$project_id,'/scatterplot?x_axis=add_counts_Avg_log10_Expr_all&y_axis=add_counts_Avg_log10_Expr_',clusterVec[i], '&cat_id=', DGEcatID,'" target="_blank">here</a>')
heatmaplink <- paste0('In order to identify cluster specific marker genes a <a href="https://biologic.crick.ac.uk/',Obio@parameterList$project_id,'/category-view/', LRcatID,'" target="_blank">cluster-maker gene heatmap for the linear-regression derrived marker gene set</a> might be interesting.')
heatmaplinkDGE <- paste0('In order to identify cluster specific marker genes a <a href="https://biologic.crick.ac.uk/',Obio@parameterList$project_id,'/category-view/', DGEcatID,'" target="_blank">cluster-maker gene heatmap for the differential-gene expression derrived marker gene set</a> might be interesting.')
figCap <- paste0(
'**Figure ',
figureCount,
':** Average gene expression intensity of all cells together versus averaged expression intensities for cluster ',clusterVec[i],'. ',
'Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>. ',
'An interactive version of this scatterplot can be found ', link, ' with the linear regression marker genes highlighted, and ',linkDGE,' with the differential gene expression marker genes highlighted. ',
heatmaplink,
heatmaplinkDGE
)
NewChnk <- paste0(
"#### ",tag,
"\n```{r LR_marker_",tag,", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",figCap,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
## Histogram Part C done ##
###########################################################################
chnkVec <- c(
chnkVec,
NewChnk
)
}
## Done creating plot array ##
###############################################################################
if (length(plotList) > 3){
tabVar <- ".tabset .tabset-fade .tabset-dropdown"
} else {
tabVar <- ".tabset .tabset-fade .tabset-pills"
}Cluster Markers By Linear Regression
## plot list will be integrated in full figure ##
cat(paste(knit(text = chnkVec, quiet = T), collapse = '\n'))LinearRegressionMarkers_ClusterC_0
Figure 24: Average gene expression intensity of all cells together versus averaged expression intensities for cluster C_0. Download a pdf of this figure here. An interactive version of this scatterplot can be found here with the linear regression marker genes highlighted, and here with the differential gene expression marker genes highlighted. In order to identify cluster specific marker genes a cluster-maker gene heatmap for the linear-regression derrived marker gene set might be interesting.In order to identify cluster specific marker genes a cluster-maker gene heatmap for the differential-gene expression derrived marker gene set might be interesting.
LinearRegressionMarkers_ClusterC_1
Figure 24: Average gene expression intensity of all cells together versus averaged expression intensities for cluster C_1. Download a pdf of this figure here. An interactive version of this scatterplot can be found here with the linear regression marker genes highlighted, and here with the differential gene expression marker genes highlighted. In order to identify cluster specific marker genes a cluster-maker gene heatmap for the linear-regression derrived marker gene set might be interesting.In order to identify cluster specific marker genes a cluster-maker gene heatmap for the differential-gene expression derrived marker gene set might be interesting.
LinearRegressionMarkers_ClusterC_2
Figure 24: Average gene expression intensity of all cells together versus averaged expression intensities for cluster C_2. Download a pdf of this figure here. An interactive version of this scatterplot can be found here with the linear regression marker genes highlighted, and here with the differential gene expression marker genes highlighted. In order to identify cluster specific marker genes a cluster-maker gene heatmap for the linear-regression derrived marker gene set might be interesting.In order to identify cluster specific marker genes a cluster-maker gene heatmap for the differential-gene expression derrived marker gene set might be interesting.
LinearRegressionMarkers_ClusterC_3
Figure 24: Average gene expression intensity of all cells together versus averaged expression intensities for cluster C_3. Download a pdf of this figure here. An interactive version of this scatterplot can be found here with the linear regression marker genes highlighted, and here with the differential gene expression marker genes highlighted. In order to identify cluster specific marker genes a cluster-maker gene heatmap for the linear-regression derrived marker gene set might be interesting.In order to identify cluster specific marker genes a cluster-maker gene heatmap for the differential-gene expression derrived marker gene set might be interesting.
LinearRegressionMarkers_ClusterC_4
Figure 24: Average gene expression intensity of all cells together versus averaged expression intensities for cluster C_4. Download a pdf of this figure here. An interactive version of this scatterplot can be found here with the linear regression marker genes highlighted, and here with the differential gene expression marker genes highlighted. In order to identify cluster specific marker genes a cluster-maker gene heatmap for the linear-regression derrived marker gene set might be interesting.In order to identify cluster specific marker genes a cluster-maker gene heatmap for the differential-gene expression derrived marker gene set might be interesting.
LinearRegressionMarkers_ClusterC_5
Figure 24: Average gene expression intensity of all cells together versus averaged expression intensities for cluster C_5. Download a pdf of this figure here. An interactive version of this scatterplot can be found here with the linear regression marker genes highlighted, and here with the differential gene expression marker genes highlighted. In order to identify cluster specific marker genes a cluster-maker gene heatmap for the linear-regression derrived marker gene set might be interesting.In order to identify cluster specific marker genes a cluster-maker gene heatmap for the differential-gene expression derrived marker gene set might be interesting.
LinearRegressionMarkers_ClusterC_6
Figure 24: Average gene expression intensity of all cells together versus averaged expression intensities for cluster C_6. Download a pdf of this figure here. An interactive version of this scatterplot can be found here with the linear regression marker genes highlighted, and here with the differential gene expression marker genes highlighted. In order to identify cluster specific marker genes a cluster-maker gene heatmap for the linear-regression derrived marker gene set might be interesting.In order to identify cluster specific marker genes a cluster-maker gene heatmap for the differential-gene expression derrived marker gene set might be interesting.
LinearRegressionMarkers_ClusterC_7
Figure 24: Average gene expression intensity of all cells together versus averaged expression intensities for cluster C_7. Download a pdf of this figure here. An interactive version of this scatterplot can be found here with the linear regression marker genes highlighted, and here with the differential gene expression marker genes highlighted. In order to identify cluster specific marker genes a cluster-maker gene heatmap for the linear-regression derrived marker gene set might be interesting.In order to identify cluster specific marker genes a cluster-maker gene heatmap for the differential-gene expression derrived marker gene set might be interesting.
LinearRegressionMarkers_ClusterC_8
Figure 24: Average gene expression intensity of all cells together versus averaged expression intensities for cluster C_8. Download a pdf of this figure here. An interactive version of this scatterplot can be found here with the linear regression marker genes highlighted, and here with the differential gene expression marker genes highlighted. In order to identify cluster specific marker genes a cluster-maker gene heatmap for the linear-regression derrived marker gene set might be interesting.In order to identify cluster specific marker genes a cluster-maker gene heatmap for the differential-gene expression derrived marker gene set might be interesting.
LinearRegressionMarkers_ClusterC_9
Figure 24: Average gene expression intensity of all cells together versus averaged expression intensities for cluster C_9. Download a pdf of this figure here. An interactive version of this scatterplot can be found here with the linear regression marker genes highlighted, and here with the differential gene expression marker genes highlighted. In order to identify cluster specific marker genes a cluster-maker gene heatmap for the linear-regression derrived marker gene set might be interesting.In order to identify cluster specific marker genes a cluster-maker gene heatmap for the differential-gene expression derrived marker gene set might be interesting.
LinearRegressionMarkers_ClusterC_10
Figure 24: Average gene expression intensity of all cells together versus averaged expression intensities for cluster C_10. Download a pdf of this figure here. An interactive version of this scatterplot can be found here with the linear regression marker genes highlighted, and here with the differential gene expression marker genes highlighted. In order to identify cluster specific marker genes a cluster-maker gene heatmap for the linear-regression derrived marker gene set might be interesting.In order to identify cluster specific marker genes a cluster-maker gene heatmap for the differential-gene expression derrived marker gene set might be interesting.
LinearRegressionMarkers_ClusterC_11
Figure 24: Average gene expression intensity of all cells together versus averaged expression intensities for cluster C_11. Download a pdf of this figure here. An interactive version of this scatterplot can be found here with the linear regression marker genes highlighted, and here with the differential gene expression marker genes highlighted. In order to identify cluster specific marker genes a cluster-maker gene heatmap for the linear-regression derrived marker gene set might be interesting.In order to identify cluster specific marker genes a cluster-maker gene heatmap for the differential-gene expression derrived marker gene set might be interesting.
LinearRegressionMarkers_ClusterC_12
Figure 24: Average gene expression intensity of all cells together versus averaged expression intensities for cluster C_12. Download a pdf of this figure here. An interactive version of this scatterplot can be found here with the linear regression marker genes highlighted, and here with the differential gene expression marker genes highlighted. In order to identify cluster specific marker genes a cluster-maker gene heatmap for the linear-regression derrived marker gene set might be interesting.In order to identify cluster specific marker genes a cluster-maker gene heatmap for the differential-gene expression derrived marker gene set might be interesting.
LinearRegressionMarkers_ClusterC_13
Figure 24: Average gene expression intensity of all cells together versus averaged expression intensities for cluster C_13. Download a pdf of this figure here. An interactive version of this scatterplot can be found here with the linear regression marker genes highlighted, and here with the differential gene expression marker genes highlighted. In order to identify cluster specific marker genes a cluster-maker gene heatmap for the linear-regression derrived marker gene set might be interesting.In order to identify cluster specific marker genes a cluster-maker gene heatmap for the differential-gene expression derrived marker gene set might be interesting.
plotList <- list()
chnkVec <- as.vector(NULL, mode = "character")
###############################################################################
## Make Heatmap ##
OsC@meta.data[["hmIdent2"]] <- paste0("C", OsC@meta.data[,Obio@parameterList$singleCellClusterString])
Idents(OsC) <- "hmIdent2"
levels <- paste0(
"C",
sort(unique(OsC@meta.data[,Obio@parameterList$singleCellClusterString]))
)
levels(OsC) <- levels
## Deal with more than 5000 cells ##
if (nrow(OsC@meta.data) > 5000){
set.seed(127)
n.cells <- 5000
OsC_HM <- OsC
OsC_HM@meta.data[["HM_sel"]] <- 0
selPos <- sample(x = nrow(OsC_HM@meta.data), size = n.cells, replace = FALSE, prob = NULL)
OsC_HM@meta.data[selPos, "HM_sel"] <- 1
OsC_HM <- subset(x = OsC_HM, subset = HM_sel == 1)
subsetString <- paste0("For this heatmap 5000 cells were randomly selected from ", nrow(OsC@meta.data)," cells in the experiment. ")
} else {
OsC_HM <- OsC
subsetString <- ""
}
## Scale Data ##
allGenes <- rownames(x = OsC_HM@assays$RNA)
OsC_HM <- ScaleData(OsC_HM, verbose = FALSE, features=allGenes)
## For the moment: resetting heatmap list to keep it short ##
Obio@parameterList[["cat2HMplotList"]] <- list()
Obio@parameterList[["cat2HMplotList"]][["Top5_Cluster_Markers"]] <- as.vector(dfTop5$gene)
## Add top transcription factors for each cluster ##
## Get transcription factor genes ##
if (Obio@parameterList$geneIDcolumn != "mgi_symbol" & Obio@parameterList$geneIDcolumn != "hgnc_symbol") {
queryGS <- "hgnc_symbol"
} else {
queryGS <- Obio@parameterList$geneIDcolumn
}
tempVec <- retrieve.gene.category.from.db(
cat_id = "ag_lab_categories__10",
password = db.pwd,
gene.symbol = queryGS,
user = Obio@parameterList$db.user,
host = Obio@parameterList$host
)
###############################################################################
## If this is fish, translation is non-human or non-mouse, translation is necessary
if (queryGS != Obio@parameterList$geneIDcolumn){
dfAnno <- Obio@dfGeneAnnotation
dfAnno <- unique(dfAnno[,c("hgnc_symbol",Obio@parameterList$geneIDcolumn )])
dfAnno <- dfAnno[dfAnno$hgnc_symbol != "", ]
dfAnno <- dfAnno[dfAnno$hgnc_symbol %in% tempVec, ]
tempVec <- unique(dfAnno[,Obio@parameterList$geneIDcolumn])
tempVec <- tempVec[tempVec != ""]
}
dfHMG <- dfGeneralMarkers[dfGeneralMarkers$gene %in% tempVec, ]
dfHMGsel <- data.frame(dfHMG %>% group_by(cluster) %>% top_n(5, avg_diff))
Obio@parameterList[["cat2HMplotList"]][["Top5_TF_Cluster_Markers"]] <- as.vector(unique(dfHMGsel$gene))
## Done with translation
for (i in 1:length(Obio@parameterList[["cat2HMplotList"]])){
tag <- paste0("HMM_", names(Obio@parameterList$cat2HMplotList)[i])
textSize <- 5
HMname <- names(Obio@parameterList[["cat2HMplotList"]])[i]
plotList[[tag]] <- DoHeatmap(
object = OsC_HM,
features = Obio@parameterList[["cat2HMplotList"]][[i]],
#group.by = "hmIdent",
draw.lines =T,
label = T,
group.bar = TRUE,
slot = "scale.data",
lines.width = 2 #With of separation lines in 'cells'
#slim.col.label = TRUE,
#remove.key = removeKey
# ) + theme(legend.position = "none"
) + theme(text = element_text(size=textSize)
) + scale_fill_gradientn(colors = c("blue", "white", "red"))
## Save to file ##
FNbase <- paste0(tag, VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
## Create R markdown chunk ##
figLegend <- paste0(
'**Figure ',
figureCount,
':** ',HMname ,' Heatmap showing the most distinct marker genes in each cluster. ' , subsetString,
'Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>. '
)
figureCount <- figureCount + 1
NewChnk <- paste0(
"#### Heatmap ", HMname,
"\n```{r Heatmap_", tag,
", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
}
rm(OsC_HM)
## Done making Heatmap ##
###############################################################################
###############################################################################
## Make dotplot
Idents(OsC) <- "hmIdent2"
levels <- paste0(
"C",
sort(unique(OsC@meta.data[,Obio@parameterList$singleCellClusterString]))
)
levels(OsC) <- levels
dpGenes <- as.vector(unique(dfTop5$gene))
if (length(dpGenes) >= 50){
dpGenes <- as.vector(unique(dfTop1$gene))
}
#dpGenes <- rev(dpGenes[!(duplicated(dpGenes))])
tag <- paste0("Dotplot_", "Var_Genes")
textSize <- 5
plotList[[tag]] <- DotPlotSB(
object = OsC,
features = dpGenes,
#cols = cols,
group.by = NULL,
split.by = NULL,
dot.scale = 4,
col.min = 0,
col.max = 5
) + ggtitle(gsub("_", "", tag)) + coord_fixed() + coord_flip()
## Save to file ##
FNbase <- paste0(tag, VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
## Create R markdown chunk ##
figLegend <- paste0(
'**Figure ',
figureCount,
':** Dotplot showing showing selected marker genes. ',
'Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>. '
)
figureCount <- figureCount + 1
NewChnk <- paste0(
"#### Dotplot Markers",
"\n```{r Dotplot_var_",
", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
## Done making dotplot ##
###############################################################################
############################
## Make cat feature plot
## done making cat feature plot
################################
if (length(plotList) > 3){
tabVar <- ".tabset .tabset-fade .tabset-dropdown"
} else {
tabVar <- ".tabset .tabset-fade .tabset-pills"
} Heatmap and Dotplot for the top5 cluster defining genes
In this section we will try to determine what are good markers for the clusters in this experiment.
cat(paste(knit(text = chnkVec, quiet = T), collapse = '\n'))Heatmap Top5_Cluster_Markers
Figure 24: Top5_Cluster_Markers Heatmap showing the most distinct marker genes in each cluster. Download a pdf of this figure here.
Heatmap Top5_TF_Cluster_Markers
Figure 25: Top5_TF_Cluster_Markers Heatmap showing the most distinct marker genes in each cluster. Download a pdf of this figure here.
Dotplot Markers
Figure 26: Dotplot showing showing selected marker genes. Download a pdf of this figure here.
library(DT)
###############################################################################
## Calculate percentages of expressed genes ##
DefaultAssay(OsC) <- "RNA"
my_genes <- rownames(x = OsC@assays$RNA)
exp <- FetchData(OsC, my_genes)
ExprMatrix <- round(as.matrix(colMeans(exp > 0)) *100,1)
colnames(ExprMatrix)[1] <- "count_cut_off"
dfExprMatrix <- data.frame(ExprMatrix)
dfExprMatrix[["gene"]] <- row.names(dfExprMatrix)
Obio@dataTableList[["dfPercCellsExpr"]] <- dfExprMatrix
## Done calculating percentages of expressed gens ##
###############################################################################
###############################################################################
## Create one table per cluster ##
## Add expressed in N percent cells ##
dfPercCellsExpr <- Obio@dataTableList$dfPercCellsExpr
dfDat <- Obio@dataTableList$dfGeneralMarkersFilt
dfDat$avg_diff <- round(dfDat$avg_diff,2)
dfDat <- dfDat[,c("cluster", "gene", "avg_diff", "power")]
names(dfDat) <- gsub("avg_diff", "DGE_avg_diff", names(dfDat))
names(dfDat) <- gsub("power", "DGE_power", names(dfDat))
dfDat[["join_col"]] <- paste0(dfDat$gene, "_C_", dfDat$cluster)
dfDat$cluster <- NULL
dfDat$gene <- NULL
## Add residual results ##
dfClusterResiduals <- Obio@dataTableList$dfClusterResiduals
library(tidyverse)
dfLongResiduals <- data.frame(
dfClusterResiduals %>% pivot_longer(!gene, names_to = "cluster", values_to = "residuals")
)
###############################################################################
## Calculate Coefficient of variation for no-cluster X genes to find best unique cluster markers
## Done ##
###############################################################################
dfLongResiduals <- dfLongResiduals[abs(dfLongResiduals$residuals) > 1, ]
dfLongResiduals$cluster <- gsub("_Residuals", "", dfLongResiduals$cluster)
dfLongResiduals[["join_col"]] <- paste0(
dfLongResiduals$gene,
"_",
dfLongResiduals$cluster
)
dfLongResiduals$gene <- NULL
dfLongResiduals$cluster <- NULL
dfDat <- merge(
dfDat,
dfLongResiduals,
by.x = "join_col",
by.y = "join_col",
all =TRUE
)
dfDat[is.na(dfDat)] <- 0
names(dfDat) <- gsub("residuals", "LinearReg_Residuals", names(dfDat))
dfDat <- dfDat[order(dfDat$LinearReg_Residuals, decreasing = T),]
dfDat[["gene"]] <- sapply(
dfDat$join_col, function(x) unlist(strsplit(x, "_"))[1]
)
dfDat[["Cluster"]] <- sapply(
dfDat$join_col, function(x) unlist(strsplit(x, "_C_"))[2]
)
dfDat$join_col <- NULL
dfDat$Cluster <- paste0("Cluster_",dfDat$Cluster, "_C")
dfDat$gene <- substr(dfDat$gene,1,50)
dfDat[["uniqueMarker"]] <- as.character(!duplicated(dfDat$gene))
dfDat$uniqueMarker <- substr(dfDat$uniqueMarker, 1,1)
dtList <- list()
tabClusters <- sort(unique(dfDat$cluster))
chnkVec <- as.vector(NULL, mode="character")
linkGeneView <- paste0('<a href="https://',urlString,'/',Obio@parameterList$project_id,'/gene-view" target = "_blank">GeneView</a>')
linkFeatureView <- paste0('<a href="https://',urlString,'/mdata/',Obio@parameterList$project_id,'/html/FeatureView.html" target="_blank">FeatureView</a>')
#for (i in 1:length(tabClusters)){
#tabLegend = paste0("**Table: ** Positive and negative marker genes for ", tabClusters[i])
tabLegend = paste0("**Table: ** Positive and negative cluster-defining marker genes. Perc_Cells_Expr: Percentage of total cells expressing gene X. Enr in Cluster: Enrichment of gene X in cluster Y. To collapse the table to one particular cluster, type the name of the cluster in the search box.",
"Use the ",linkGeneView," or ",linkFeatureView," functionalities to examine individual genes in more detail. "
)
#dfTempDat <- dfDat[dfDat$cluster == tabClusters[i],]
dfTempDat <- dfDat
## Percent expressed genes
dfTempDat <- merge(
dfTempDat,
Obio@dataTableList$dfPercCellsExpr,
by.x = "gene",
by.y = "gene"
)
names(dfTempDat) <- gsub("count_cut_off", "Perc_Cells_Expr",names(dfTempDat))
names(dfTempDat) <- gsub("myAUC", "AUC", names(dfTempDat))
names(dfTempDat) <- gsub("[.]", "", names(dfTempDat))
#dtList[[paste0("Table",i)]] <- datatable(dfDat,rownames = FALSE)
if (Obio@parameterList$host == "10.27.241.234"){
urlString <- "biologic.thecrick.org"
} else {
urlString <- "biologic.crick.ac.uk"
}
dfTempDat$gene <- paste0('<a href="https://',urlString,'/',Obio@parameterList$project_id,'/gene-view?query=',dfTempDat$gene,'&exact=TRUE" target="_blank">', dfTempDat$gene, '</a>')
NewChnk <- paste0(
"#### ", names(dtList),
"\n```{r datatable_",
i,", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
tabLegend,"'}\n",
"\n",
"\n datatable(dfTempDat,rownames = FALSE, escape = FALSE)",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
#}
## Done creating one table per cluster ##
##############################################################################Cluster-defining Genes Table
cat(paste(knit(text = chnkVec, quiet = T), collapse = '\n'))Table: Positive and negative cluster-defining marker genes. Perc_Cells_Expr: Percentage of total cells expressing gene X. Enr in Cluster: Enrichment of gene X in cluster Y. To collapse the table to one particular cluster, type the name of the cluster in the search box.Use the GeneView or FeatureView functionalities to examine individual genes in more detail.
###########################
## Create gmt list
## Retrieve gmt files from database
## Add custom gmt files
library(clusterProfiler)
library(ggplot2)
gmtList <- list()
dbtableList <- list(
"Cell Type Signatures" = "mysigdb_sc_sig",
"Cell Type Signatures" = "cibersort_L22",
"GO-MF" = "mysigdb_c5_MF",
"Pathways" = "mysigdb_c2_1329_canonical_pathways"
)
for (i in 1:length(dbtableList)){
dfTemp <- unique(import.db.table.from.db(
host = Obio@dbDetailList$host,
dbname = Obio@dbDetailList$ref.cat.db,
dbtable = dbtableList [[i]],
password = db.pwd,
user = Obio@dbDetailList$db.user
))
rmVec <- grep("temp_", dfTemp$cat_type)
if (length(rmVec) > 0){
dfTemp <- dfTemp[-rmVec, ]
}
dfTemp <- unique(dbcat2gmt(
dfTemp, # As downloaded from reference_categories_db_new database
gene.id.column = queryGS
))
dfTemp <- dfTemp[!duplicated(dfTemp[,1]), ]
write.table(
dfTemp,
"temp.gmt.txt",
row.names = F,
sep = "\t",
col.names = F,
quote = F
)
CPgmt <- read.gmt("temp.gmt.txt")
unlink("temp.gmt.txt")
CPgmt <- unique(CPgmt[CPgmt$gene != "", ])
gmtList[[dbtableList[[i]]]] <- CPgmt
}
## Edit collection names for plot
names(gmtList) <- gsub("mysigdb_", "",names(gmtList))
names(gmtList) <- gsub("c2_1329_canonical_p", "P",names(gmtList))
names(gmtList) <- gsub("sc_sig", "CellSig",names(gmtList))
names(gmtList) <- gsub("cibersort_L22", "CellSig",names(gmtList))
names(gmtList) <- gsub("c5_", "GO_",names(gmtList))
## Done creating gmt list
###########################
resList <- profileCluster(
# Input gmt file with categories to test: dfGmt
# Output: table with enrichments
obj = Obio,
markerList = Obio@dataTableList[["residualClusterMarkers"]],
gmtList = gmtList,
nTop = 10,
pvalueCutoff= 0.5
)
plotList <- resList$plotList
chnkVec <- resList$chnkVec
if (length(plotList) > 3){
tabVar <- ".tabset .tabset-fade .tabset-dropdown"
} else {
tabVar <- ".tabset .tabset-fade .tabset-pills"
}Define Cell Type Categories Enriched in Individual Clusters
cat(paste(knit(text = chnkVec, quiet = T), collapse = '\n'))Cell Types C 0 Residuals Linear Regression Markers 1
Figure 27: Enrichment analysis of cluster gene signatures to infer the cluster cell type. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > Cell Signatures and find these categories using the search box.
Cell Types C 1 Residuals Linear Regression Markers 1
Figure 28: Enrichment analysis of cluster gene signatures to infer the cluster cell type. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > Cell Signatures and find these categories using the search box.
Cell Types C 10 Residuals Linear Regression Markers 1
Figure 29: Enrichment analysis of cluster gene signatures to infer the cluster cell type. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > Cell Signatures and find these categories using the search box.
Cell Types C 11 Residuals Linear Regression Markers 1
Figure 30: Enrichment analysis of cluster gene signatures to infer the cluster cell type. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > Cell Signatures and find these categories using the search box.
Cell Types C 12 Residuals Linear Regression Markers 1
Figure 31: Enrichment analysis of cluster gene signatures to infer the cluster cell type. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > Cell Signatures and find these categories using the search box.
Cell Types C 13 Residuals Linear Regression Markers 1
Figure 32: Enrichment analysis of cluster gene signatures to infer the cluster cell type. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > Cell Signatures and find these categories using the search box.
Cell Types C 2 Residuals Linear Regression Markers 1
Figure 33: Enrichment analysis of cluster gene signatures to infer the cluster cell type. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > Cell Signatures and find these categories using the search box.
Cell Types C 3 Residuals Linear Regression Markers 1
Figure 34: Enrichment analysis of cluster gene signatures to infer the cluster cell type. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > Cell Signatures and find these categories using the search box.
Cell Types C 4 Residuals Linear Regression Markers 1
Figure 35: Enrichment analysis of cluster gene signatures to infer the cluster cell type. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > Cell Signatures and find these categories using the search box.
Cell Types C 5 Residuals Linear Regression Markers 1
Figure 36: Enrichment analysis of cluster gene signatures to infer the cluster cell type. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > Cell Signatures and find these categories using the search box.
Cell Types C 6 Residuals Linear Regression Markers 1
Figure 37: Enrichment analysis of cluster gene signatures to infer the cluster cell type. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > Cell Signatures and find these categories using the search box.
Cell Types C 7 Residuals Linear Regression Markers 1
Figure 38: Enrichment analysis of cluster gene signatures to infer the cluster cell type. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > Cell Signatures and find these categories using the search box.
Cell Types C 8 Residuals Linear Regression Markers 1
Figure 39: Enrichment analysis of cluster gene signatures to infer the cluster cell type. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > Cell Signatures and find these categories using the search box.
Cell Types C 9 Residuals Linear Regression Markers 1
Figure 40: Enrichment analysis of cluster gene signatures to infer the cluster cell type. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > Cell Signatures and find these categories using the search box.
Gene Set Heatmaps
## Two options: full heatmap and averaged heatmaps
## https://satijalab.org/seurat/v3.0/interaction_vignette.html
###############################################################################
## Add percentage expressed genes ##
# DefaultAssay(OsC) <- "RNA"
# my_genes <- rownames(x = OsC@assays$RNA)
#
# exp <- FetchData(OsC, my_genes)
#
# ExprMatrix <- round(as.matrix(colMeans(exp > 0)) *100,1)
# colnames(ExprMatrix)[1] <- "count_cut_off"
# dfExprMatrix <- data.frame(ExprMatrix)
# dfExprMatrix[["gene"]] <- row.names(dfExprMatrix)
#
# Obio@dataTableList[["dfPercCellsExpr"]] <- dfExprMatrix
#
# hmRelevantGenes <- as.vector(unique(dfExprMatrix[dfExprMatrix$count_cut_off > Obio@parameterList$singleCellPercExpressedMinCutOff, "gene"]))
#
#
#
# ## Done adding percentage expressed ##
# ###############################################################################
#
#
# ###############################################################################
# ## Make plot according to reference categories ##
# allGenes <- rownames(x = OsC@assays$RNA)
# OsC <- ScaleData(OsC, verbose = FALSE, features=allGenes)
#
# DefaultAssay(OsC) <- "RNA"
#
#
# ## Add heatmap identities to meta.data ##
#
OsC@meta.data[["hmIdent"]] <- paste0(
OsC@meta.data[,Obio@parameterList$singleCellClusterString],
"_",
substr(OsC@meta.data$sampleID,1,10)
)
#
if (length(unique(OsC@meta.data$hmIdent)) > 25){
OsC@meta.data[["hmIdent"]] <- OsC@meta.data[,Obio@parameterList$singleCellClusterString]
}
#
# Idents(OsC) <- "hmIdent"
#
# ## Done adding heatmap identities to meta.data ##
#
# printPdf <- TRUE
# referenceList <- Obio@dataTableList$referenceList
# for (i in 1:length(referenceList)){
# HMname <- names(referenceList)[i]
# geneVec <- unique(referenceList[[i]][referenceList[[i]] %in% rownames(x = OsC@assays$RNA)])
#
# if (length(geneVec) > 50){
# geneVec <- geneVec[geneVec %in% hmRelevantGenes]
# }
#
#
# ## Do Heatmap ##
# if (length(geneVec) < 1500 & length(geneVec) > 2){
# Idents(OsC) <- "hmIdent"
# HMname <- names(referenceList)[i]
# cat("\n")
# cat(paste0("**Heatmap ", HMname,"**"))
# cat("\n")
# cat("\n")
# ## Cluster genes ##
# HMgenes <- referenceList[[i]]
# #dfCluster <- OsC@assays$integrated
# Mexpr <- GetAssayData(object = OsC, assay.type = "integrated", slot = "scale.data")
# HMgenesSel <- HMgenes[HMgenes %in% row.names(Mexpr)]
#
# if (length(HMgenesSel) > 2){
# Mexpr <- Mexpr[HMgenesSel,]
#
# pdf(paste0("temp", VersionPdfExt))
# hmRes <- make.hm(
# m.df1 = Mexpr,
# filename = "",
# k.number = 1,
# n.colors = 1000,
# hclust.method = "complete",
# dist.method = "euclidean",
# main = "",
# Colv = TRUE,
# showRowNames = TRUE,
# showColNames = F,
# plotSeparationLines = FALSE
# )
# dev.off()
#
# orderedGenes <- as.vector(unique(row.names(hmRes$sorted)))
#
# if (length(unique(OsC@meta.data$hmIdent)) > 10){
# removeKey <- TRUE
# } else {
# removeKey <- FALSE
# }
#
# if (length(orderedGenes) <= 50){
# label = TRUE
# } else {
# label = FALSE
# }
#
# p1 <- DoHeatmap(
# object = OsC,
# features = orderedGenes,
# #group.by = "hmIdent",
# draw.lines =T,
# label = label,
# group.bar = TRUE,
# slot = "scale.data",
# lines.width = 2 #With of separation lines in 'cells'
# #slim.col.label = TRUE,
# #remove.key = removeKey
# ) + theme(legend.position = "none")
#
# print(p1)
#
# ## Save to file ##
# FNbase <- paste0("HM", HMname, VersionPdfExt)
# FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
# FNrel <- paste0("report_figures/", FNbase)
#
# pdf(FN)
# print(p1)
# dev.off()
#
# cat("\n")
# cat(paste0('Download a pdf of this figure [here](', FNrel, '). '))
# cat("\n")
# cat("\n")
#
# }
# }
# }
## Done making plots according to gene categories ##
###############################################################################Example Feature Plots
DefaultAssay(OsC) <- "RNA"
cat(paste0(
'Feature plots for any gene in this experiment can be viewed <a href="https://biologic.crick.ac.uk/mdata/',Obio@parameterList$project_id,'/html/FeatureView.html" target="_blank">here</a>.'
))Feature plots for any gene in this experiment can be viewed here.
plotGenes <- c(
Obio@dataTableList$referenceList$integrated_top30var[1:10]
)
plotParts <- ceiling(length(plotGenes)/2)
chnkVec <- as.vector(NULL, mode = "character")
plotListF <- list()
for (i in 1:plotParts){
tag1 <- paste0("Featureplot_",i)
featureGenes <- c(plotGenes[((2*i)-1)], plotGenes[((2*i))])
plotListF[[tag1]] <- FeaturePlot(
OsC,
features = featureGenes,
#split.by = "orig.ident",
reduction = Obio@parameterList$primReduction
)
## Save to file ##
FNbase <- paste0("Featureplot.", plotGenes[((2*i)-1)], ".", plotGenes[((2*i))],".", VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotListF[[tag1]])
dev.off()
linkFeatureView <- paste0('<a href="https://',urlString,'/mdata/',Obio@parameterList$project_id,'/html/FeatureView.html" target="_blank">FeatureView</a>')
figLegend <- paste0(
"**Figure ",
figureCount,
":** Gene expression plot for genes ",
plotGenes[((2*i)-1)],
" and ",
plotGenes[((2*i))],".",
" Results for any other gene may be plotted in ",linkFeatureView,"."
)
NewChnk <- paste0(
"#### Featureplot ", plotGenes[((2*i)-1)], " and ",plotGenes[((2*i))],
"\n```{r FeaturePlot_", i,
", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotListF[['",tag1,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
}cat(paste(knit(text = chnkVec, quiet = T), collapse = '\n'))Featureplot Sst and Pcdh11x
Figure 27: Gene expression plot for genes Sst and Pcdh11x. Results for any other gene may be plotted in FeatureView.
Featureplot Vip and Klhl1
Figure 27: Gene expression plot for genes Vip and Klhl1. Results for any other gene may be plotted in FeatureView.
Featureplot Gal and Penk
Figure 27: Gene expression plot for genes Gal and Penk. Results for any other gene may be plotted in FeatureView.
Featureplot Tac1 and RP23-231J2.1
Figure 27: Gene expression plot for genes Tac1 and RP23-231J2.1. Results for any other gene may be plotted in FeatureView.
Featureplot Calcb and Pdyn
Figure 27: Gene expression plot for genes Calcb and Pdyn. Results for any other gene may be plotted in FeatureView.
PCA Gene and Cell Plots
figureCount <- figureCount + 1
# An alternative heuristic method generates an 'Elbow plot': a ranking of principle components based on the percentage of variance explained by each one (`ElbowPlot` function). In this example, we can observe an 'elbow' around PC9-10, suggesting that the majority of true signal is captured in the first 10 PCs.
ElbowPlot(object = OsC) + theme(
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12)
) + ggtitle(paste0("Variance per PCA Dimension"))
Figure, 28: Elbowplot Variance explained per PCA dimension
## Plot variance per PCA dimension #### Add PCA coordinates ##
dfTemp <- data.frame(OsC@reductions$pca@cell.embeddings)[, 1:20]
OsC <- addDf2seuratMetaData(
obj = OsC,
dfAdd = dfTemp
)
## Add UMAP coordinates to Metadata ##
dfAdd <- data.frame(OsC@reductions$umap@cell.embeddings)
OsC <- addDf2seuratMetaData(
obj = OsC,
dfAdd = dfAdd
)
## Add tSNE coordinates to Metadata ##
dfAdd <- data.frame(OsC@reductions$tsne@cell.embeddings)
OsC <- addDf2seuratMetaData(
obj = OsC,
dfAdd = dfAdd
)
Obio@dataTableList[["meta.data"]] <- OsC@meta.data
xVec <- c("PC_1","PC_3","PC_5","PC_7","PC_9","PC_11","PC_13","PC_15","PC_17","PC_19")
yVec <- c("PC_2","PC_4","PC_6","PC_8","PC_10","PC_12","PC_14","PC_16","PC_18","PC_20")
pcVec <- c("PC_1","PC_2","PC_3","PC_4","PC_5","PC_6","PC_7","PC_8","PC_9","PC_10")
chnkVec <- as.vector(NULL, mode = "character")
plotListCell <- list()
plotListGene <- list()
###############################################################################
## Collect top-enriched genes ##
EnrichedGenesList <- list()
## Done ##
###############################################################################
for (i in 1:length(xVec)){
dfDat <- Obio@dataTableList$meta.data
dfSel <- dfDat
selXY <- c(xVec[i], yVec[i])
colCol <- Obio@parameterList$singleCellClusterString
tag <- paste0(xVec[i], "and", yVec[i])
tag <- gsub("_", "", tag)
## Make Cell level PCA
plotListCell[[tag]] <- ggplot(data=dfDat, aes_string(selXY[1] , selXY[2], col=colCol, shape="sampleID")
) + geom_vline(xintercept = 0, color = "grey", size=0.5
) + geom_hline(yintercept = 0, color = "grey", size=0.5
) + geom_point()+ ggtitle(paste0("PCA - Cell Level")
) + theme_bw(
) + theme(
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12)
)
## Save to file ##
FNbase <- paste0("PCA.cell.level.", xVec[i],".", yVec[i], ".", VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotListCell[[tag]])
dev.off()
link <- paste0('<a href="https://',urlString,'/',Obio@parameterList$project_id,'/pca?x_axis=',gsub('_', '', xVec[i]),'&y_axis=',gsub('_', '', yVec[i]),'" target = "_blank">here</a>')
figCap <- paste0(
'**Figure, ' ,figureCount,'A:** Cell-level PCA plot for dimensions ', xVec[i], ' and ', yVec[i],'. ',
'Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>. ',
'An interactive version of this figure can be found ', link, '. '
)
NewChnk <- paste0(
"#### PCA Cell Level ", xVec[i], " and ",yVec[i],
"\n```{r PCAcells_", i,
", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figCap,"'}\n",
"\n",
"\n print(plotListCell[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
## Done with cells ##
###########################################################################
###########################################################################
## Start with genes ##
dfPCADat <- data.frame(Loadings(OsC, reduction = "pca"))
dfPCADat[["gene"]] <- row.names(dfPCADat)
dfPCADat <- gather(
dfPCADat,
condition,
measurement, 1:(ncol(dfPCADat)-1),
factor_key=TRUE
)
Obio@dataTableList[["dfPCAloadings"]] <- dfPCADat
## Make Gene Level PCA ##
dfPCADat <- data.frame(Loadings(OsC, reduction = "pca"))
dfPCADat[["gene"]] <- row.names(dfPCADat)
dfPCADat <- gather(
dfPCADat,
condition,
measurement, 1:(ncol(dfPCADat)-1),
factor_key=TRUE
)
dfLoad <- dfPCADat
Obio@dataTableList$dfPCAloadings <- dfLoad
## Plot ##
selXY <- c(xVec[i], yVec[i])
dfSel <- filter(dfLoad, condition %in% selXY)
dfSel <- dfSel %>% spread(key=condition, value=measurement)
row.names(dfSel) <- dfSel$gene
dfSel[["highlight"]] <- ""
dfSel <- dfSel[order(dfSel[,selXY[1]], decreasing = FALSE), ]
dfSel[1:15, "highlight"] <- "+"
EnrichedGenesList[[paste0(selXY[1], "_neg")]]<- as.vector(dfSel$gene[1:15])
dfSel <- dfSel[order(dfSel[,selXY[1]], decreasing = TRUE), ]
dfSel[1:15, "highlight"] <- "+"
EnrichedGenesList[[paste0(selXY[1], "_pos")]]<- as.vector(dfSel$gene[1:15])
dfSel <- dfSel[order(dfSel[,selXY[2]], decreasing = FALSE), ]
dfSel[1:15, "highlight"] <- "+"
EnrichedGenesList[[paste0(selXY[2], "_neg")]]<- as.vector(dfSel$gene[1:15])
dfSel <- dfSel[order(dfSel[,selXY[2]], decreasing = TRUE), ]
dfSel[1:15, "highlight"] <- "+"
EnrichedGenesList[[paste0(selXY[2], "_pos")]]<- as.vector(dfSel$gene[1:15])
plotListGene[[tag]] <- ggplot(data=dfSel, aes_string(x=selXY[1],y=selXY[2], col="highlight")
) + geom_vline(xintercept = 0, color = "grey", size=0.5
) + geom_hline(yintercept = 0, color = "grey", size=0.5) + geom_point() + scale_color_manual(values=c("black", "red")) + ggtitle(paste0("PCA - Cell Level"
)) + theme_bw(
) + theme(
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12)
)
points <- as.vector(unique(dfSel[dfSel$highlight=="+", "gene"]))
plotListGene[[tag]] <- LabelPoints(plot = plotListGene[[tag]], points =points, repel = TRUE, xnudge = 0, ynudge = 0)
## Save to file ##
FNbase <- paste0("PCA.cell.level.", xVec[i],".", yVec[i],".", VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotListGene[[tag]])
dev.off()
dim1 <- gsub("PC_", "", xVec[i])
dim2 <- gsub("PC_", "", yVec[i])
link <- paste0(
'<a href="https://',urlString,'/',
Obio@parameterList$project_id,
'/scatterplot?x_axis=add_counts_PCA_Dim_',
dim1,
'_Loadings&y_axis=add_counts_PCA_Dim_',
dim2,
'_Loadings&highlight_gene=&cat_id=ag_lab_categories__10',
'" target="_blank">here</a>'
)
figCap <- paste0(
"**Figure, " ,figureCount,"B:**Gene-level PCA plot for dimensions ", xVec[i], " and ", yVec[i], ". An interactive version of this figure can be found ", link, ". "
)
NewChnk <- paste0(
"\n```{r PCA_gene_level_", i,
", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figCap,"'}\n",
"\n",
"\n print(plotListGene[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
## Done with genes ##
###########################################################################
figureCount <- figureCount + 1
}
if (length(plotListGene) > 3){
tabVar <- ".tabset .tabset-fade .tabset-dropdown"
} else {
tabVar <- ".tabset .tabset-fade .tabset-pills"
}Characterization of PCA Loadings
cat(paste(knit(text = chnkVec, quiet = T), collapse = '\n'))PCA Cell Level PC_1 and PC_2
Figure, 28A: Cell-level PCA plot for dimensions PC_1 and PC_2. Download a pdf of this figure here. An interactive version of this figure can be found here.
Figure, 28B:Gene-level PCA plot for dimensions PC_1 and PC_2. An interactive version of this figure can be found here.
PCA Cell Level PC_3 and PC_4
Figure, 29A: Cell-level PCA plot for dimensions PC_3 and PC_4. Download a pdf of this figure here. An interactive version of this figure can be found here.
Figure, 29B:Gene-level PCA plot for dimensions PC_3 and PC_4. An interactive version of this figure can be found here.
PCA Cell Level PC_5 and PC_6
Figure, 30A: Cell-level PCA plot for dimensions PC_5 and PC_6. Download a pdf of this figure here. An interactive version of this figure can be found here.
Figure, 30B:Gene-level PCA plot for dimensions PC_5 and PC_6. An interactive version of this figure can be found here.
PCA Cell Level PC_7 and PC_8
Figure, 31A: Cell-level PCA plot for dimensions PC_7 and PC_8. Download a pdf of this figure here. An interactive version of this figure can be found here.
Figure, 31B:Gene-level PCA plot for dimensions PC_7 and PC_8. An interactive version of this figure can be found here.
PCA Cell Level PC_9 and PC_10
Figure, 32A: Cell-level PCA plot for dimensions PC_9 and PC_10. Download a pdf of this figure here. An interactive version of this figure can be found here.
Figure, 32B:Gene-level PCA plot for dimensions PC_9 and PC_10. An interactive version of this figure can be found here.
PCA Cell Level PC_11 and PC_12
Figure, 33A: Cell-level PCA plot for dimensions PC_11 and PC_12. Download a pdf of this figure here. An interactive version of this figure can be found here.
Figure, 33B:Gene-level PCA plot for dimensions PC_11 and PC_12. An interactive version of this figure can be found here.
PCA Cell Level PC_13 and PC_14
Figure, 34A: Cell-level PCA plot for dimensions PC_13 and PC_14. Download a pdf of this figure here. An interactive version of this figure can be found here.
Figure, 34B:Gene-level PCA plot for dimensions PC_13 and PC_14. An interactive version of this figure can be found here.
PCA Cell Level PC_15 and PC_16
Figure, 35A: Cell-level PCA plot for dimensions PC_15 and PC_16. Download a pdf of this figure here. An interactive version of this figure can be found here.
Figure, 35B:Gene-level PCA plot for dimensions PC_15 and PC_16. An interactive version of this figure can be found here.
PCA Cell Level PC_17 and PC_18
Figure, 36A: Cell-level PCA plot for dimensions PC_17 and PC_18. Download a pdf of this figure here. An interactive version of this figure can be found here.
Figure, 36B:Gene-level PCA plot for dimensions PC_17 and PC_18. An interactive version of this figure can be found here.
PCA Cell Level PC_19 and PC_20
Figure, 37A: Cell-level PCA plot for dimensions PC_19 and PC_20. Download a pdf of this figure here. An interactive version of this figure can be found here.
Figure, 37B:Gene-level PCA plot for dimensions PC_19 and PC_20. An interactive version of this figure can be found here.
selVec <- c(Obio@parameterList$singleCellClusterString, names(OsC@meta.data)[grep("^PC", names(OsC@meta.data))])
dfPCdat <- OsC@meta.data[, selVec]
ymin <- 1.1*min(dfPCdat$PC_1)
ymax <- 1.1*max(dfPCdat$PC_1)
clusterVec <- sort(unique(OsC@meta.data[,Obio@parameterList$singleCellClusterString]))
library(scales)
clusterColVec <- hue_pal()(length(clusterVec))
i=1
chnkVec <- as.vector(NULL, mode="character")
plotList <- list()
for (i in 1:length(clusterVec)){
dfTemp <- dfPCdat[dfPCdat[,Obio@parameterList$singleCellClusterString] == clusterVec[i],]
dfTemp[,Obio@parameterList$singleCellClusterString] <- NULL
library(tidyr)
dfTemp <- gather(dfTemp, PC)
orderVec <- sort(as.numeric(gsub("PC_", "",unique(dfTemp$PC))))
orderVec <- paste0("PC_", orderVec)
dfTemp$PC <- factor(dfTemp$PC, levels = orderVec)
Ncolumns <- length(unique(dfTemp$PC))
a <- paste0("Cluster_", clusterVec[i])
tag <- paste0("PCA_Distributions_", a)
plotList[[tag]] <-ggplot(
dfTemp,
aes(x=PC, y=value, fill = PC)
) + geom_hline(yintercept = 0, color = "black", size=0.5
) + geom_jitter(width=0.1,alpha=0.2
) + geom_boxplot(
) + theme_bw(
) + theme(
legend.position = "none",
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8, angle = 45,vjust = 1, hjust=1),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12)
) + ggtitle(paste0("PCA Distribution: ", a)
) + ylim(ymin, ymax) + scale_fill_manual(values=rep(clusterColVec[i], Ncolumns))
## Save to file ##
FNbase <- paste0(tag, VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
figCap <- paste0(
'**Figure, ' ,
figureCount,
':** This plot may help you to identify PCA dimensions, in which marker genes for cluster ',
clusterVec[i],
' become evident. Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>. '
)
figureCount <- figureCount + 1
NewChnk <- paste0(
"#### ", tag,
"\n```{r ", tag,
", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figCap,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
}
if (length(plotList) > 3){
tabVar <- ".tabset .tabset-fade .tabset-dropdown"
} else {
tabVar <- ".tabset .tabset-fade .tabset-pills"
}Characterize PCA Dimensions
cat(paste(knit(text = chnkVec, quiet = T), collapse = '\n'))PCA_Distributions_Cluster_0
Figure, 38: This plot may help you to identify PCA dimensions, in which marker genes for cluster 0 become evident. Download a pdf of this figure here.
PCA_Distributions_Cluster_1
Figure, 39: This plot may help you to identify PCA dimensions, in which marker genes for cluster 1 become evident. Download a pdf of this figure here.
PCA_Distributions_Cluster_2
Figure, 40: This plot may help you to identify PCA dimensions, in which marker genes for cluster 2 become evident. Download a pdf of this figure here.
PCA_Distributions_Cluster_3
Figure, 41: This plot may help you to identify PCA dimensions, in which marker genes for cluster 3 become evident. Download a pdf of this figure here.
PCA_Distributions_Cluster_4
Figure, 42: This plot may help you to identify PCA dimensions, in which marker genes for cluster 4 become evident. Download a pdf of this figure here.
PCA_Distributions_Cluster_5
Figure, 43: This plot may help you to identify PCA dimensions, in which marker genes for cluster 5 become evident. Download a pdf of this figure here.
PCA_Distributions_Cluster_6
Figure, 44: This plot may help you to identify PCA dimensions, in which marker genes for cluster 6 become evident. Download a pdf of this figure here.
PCA_Distributions_Cluster_7
Figure, 45: This plot may help you to identify PCA dimensions, in which marker genes for cluster 7 become evident. Download a pdf of this figure here.
PCA_Distributions_Cluster_8
Figure, 46: This plot may help you to identify PCA dimensions, in which marker genes for cluster 8 become evident. Download a pdf of this figure here.
PCA_Distributions_Cluster_9
Figure, 47: This plot may help you to identify PCA dimensions, in which marker genes for cluster 9 become evident. Download a pdf of this figure here.
PCA_Distributions_Cluster_10
Figure, 48: This plot may help you to identify PCA dimensions, in which marker genes for cluster 10 become evident. Download a pdf of this figure here.
PCA_Distributions_Cluster_11
Figure, 49: This plot may help you to identify PCA dimensions, in which marker genes for cluster 11 become evident. Download a pdf of this figure here.
PCA_Distributions_Cluster_12
Figure, 50: This plot may help you to identify PCA dimensions, in which marker genes for cluster 12 become evident. Download a pdf of this figure here.
PCA_Distributions_Cluster_13
Figure, 51: This plot may help you to identify PCA dimensions, in which marker genes for cluster 13 become evident. Download a pdf of this figure here.
library(knitr)
library(ggplot2)
#save.image("temp.RData")
PCAdimensions <- paste0("PC_", 1:20)
plotList <- list()
chnkVec <- as.vector(NULL, mode = "character")
for (j in 1:length(PCAdimensions)){
posTestGeneSet <- as.vector(
unique(
EnrichedGenesList[[paste0(PCAdimensions[j], "_pos")]]
)
)
negTestGeneSet <- as.vector(
unique(
EnrichedGenesList[[paste0(PCAdimensions[j], "_neg")]]
)
)
## Get background gene set ##
#backgroundGeneVec <- row.names(OsC[["RNA"]]@counts)
if ((length(posTestGeneSet) >= 3) |(length(negTestGeneSet) >= 3)){
library(enrichR)
topMaxCat <- 10
dbs <- listEnrichrDbs()
dbs <- c("GO_Biological_Process_2017")
PosEnriched <- enrichr(posTestGeneSet, dbs)
for (i in 1:length(dbs)){
dfTemp <- PosEnriched[[dbs[i]]]
if (i ==1){
dfPosEnriched <- dfTemp
} else {
dfPosEnriched <- rbind(
dfPosEnriched,
dfTemp
)
}
}
dfPosEnriched[["log10FDR"]] <- -1*log10(dfPosEnriched$Adjusted.P.value)
dfPosEnriched <- dfPosEnriched[order(-dfPosEnriched$log10FDR),]
dfPosEnriched <- na.omit(dfPosEnriched)
## Negative Side ##
NegEnriched <- enrichr(negTestGeneSet, dbs)
for (i in 1:length(dbs)){
dfTemp <- NegEnriched[[dbs[i]]]
if (i ==1){
dfNegEnriched <- dfTemp
} else {
dfNegEnriched <- rbind(
dfNegEnriched,
dfTemp
)
}
}
dfNegEnriched[["log10FDR"]] <- -1*log10(dfNegEnriched$Adjusted.P.value)
dfNegEnriched <- dfNegEnriched[order(-dfPosEnriched$log10FDR),]
dfNegEnriched <- na.omit(dfNegEnriched)
dfNegSel <- dfNegEnriched
if (nrow(dfNegSel) > topMaxCat){
dfNegSel <- dfNegSel[1:topMaxCat,]
}
dfPosSel <- dfPosEnriched
if (nrow(dfPosSel) > topMaxCat){
dfPosSel <- dfPosSel[1:topMaxCat,]
}
if ((nrow(dfNegEnriched) > 0) | (nrow(dfPosEnriched) > 0)){
dfNegSel$log10FDR <- -1* dfNegSel$log10FDR
dfSel <- rbind(
dfNegSel,
dfPosSel
)
dfSel <- na.omit(dfSel)
dfSel <- dfSel[order(dfSel$log10FDR),]
dfSel$log10FDR <- round(dfSel$log10FDR, 2)
dfSel[["Category"]] <- ""
dfSel[dfSel$log10FDR >= 0, "Category"] <- "Enr."
dfSel[dfSel$log10FDR < 0, "Category"] <- "Depl."
for (k in 1:nrow(dfSel)){
if (nchar(dfSel[k, "Term"]) > 50 & length(grep("\\(GO", as.vector(dfSel[k, "Term"]))) > 0){
part1 <- unlist(strsplit(as.vector(dfSel[k, "Term"]), "\\(GO"))[1]
part1 <- substr(part1, 1, 45)
part2 <- unlist(strsplit(as.vector(dfSel[k, "Term"]), "\\(GO"))[2]
part2 <- paste0("\\(GO", part2)
if (nchar(part1) > 40 ){
dfSel[k, "Term"] <- paste0(part1, " \\n", part2)
} else {
dfSel[k, "Term"] <- paste0(part1, " ", part2)
}
}
}
#dfSel$Term <- gsub("\\(GO", "\\\n\\(GO", dfSel$Term)
dfSel$Term <- factor(dfSel$Term, levels = unique(dfSel$Term))
plotList[[paste0("PCA_ENR_", j)]] <- ggplot(
data=dfSel, aes(x= Term, y=log10FDR, fill=Category, order=log10FDR)
) + geom_bar(stat="identity", colour="black"
) + coord_flip() +scale_fill_manual(values=c("yellow", "blue")
) + theme_bw(
) + theme(
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12)
) + labs(title = paste0("Cluster ", PCAdimensions[j]," enriched genes") ,y = "-log10(FDR)", x = ""
) + geom_hline(yintercept = c(-log10(0.05), log10(0.05)), color = "red", size=0.5, lty=2
) + geom_hline(yintercept = 0, color = "black", size=0.5
)
cat(" \n")
## Save to file ##
FNbase <- paste0("PCA_Cluster_", PCAdimensions[j],".enriched.genes", VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[paste0("PCA_ENR_", j)]])
dev.off()
link <- paste0(
'<a href="https://', urlString, '/',
Obio@parameterList$project_id,
'/category-view?category_type=GO-BP" target="_blank">CategoryView > GO-BP</a>'
)
## Create R markdown chunk ##
figLegend <- paste0(
'**Figure ',
figureCount,
'**: GO-BP category enrichment analysis for the 15 genes that have <font color = "yellow"> the most positive </font> and <font color = "blue">the most negative</font> PCA loading values in dimension ',
PCAdimensions[j],
' associated with them. Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>. To view these gene sets in the context of your data, go to ',link,' and find these categories using the search box.'
)
figureCount <- figureCount + 1
NewChnk <- paste0(
"#### ", PCAdimensions[j],
"\n```{r enrichr_",
j,", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",paste0("PCA_ENR_", j),"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
}
}
chnkVec <- c(
chnkVec,
NewChnk
)
}
if (length(plotList) > 3){
tabVar <- ".tabset .tabset-fade .tabset-dropdown"
} else {
tabVar <- ".tabset .tabset-fade .tabset-pills"
}Characterize PCA Dimensions
###############################################################################
## Do category enrichment on clusters ##
cat(paste(knit(text = chnkVec, quiet = T), collapse = '\n'))PC_1
Figure 52: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_1 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
PC_2
Figure 53: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_2 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
PC_3
Figure 54: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_3 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
PC_4
Figure 55: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_4 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
PC_5
Figure 56: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_5 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
PC_6
Figure 57: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_6 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
PC_7
Figure 58: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_7 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
PC_8
Figure 59: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_8 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
PC_9
Figure 60: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_9 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
PC_10
Figure 61: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_10 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
PC_11
Figure 62: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_11 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
PC_12
Figure 63: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_12 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
PC_13
Figure 64: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_13 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
PC_14
Figure 65: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_14 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
PC_15
Figure 66: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_15 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
PC_16
Figure 67: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_16 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
PC_17
Figure 68: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_17 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
PC_18
Figure 69: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_18 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
PC_19
Figure 70: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_19 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
PC_20
Figure 71: GO-BP category enrichment analysis for the 15 genes that have the most positive and the most negative PCA loading values in dimension PC_20 associated with them. Download a pdf of this figure here. To view these gene sets in the context of your data, go to CategoryView > GO-BP and find these categories using the search box.
## Done doing enrichment on clusters ##
################################################################################# Add default ##
pos <- grep("addDmaps", names(Obio@parameterList))
if (length(pos) == 0){
Obio@parameterList[["addDmaps"]] <- FALSE
}
if (Obio@parameterList$addDmaps){
library(knitr)
library(ggplot2)
invertPT <- TRUE
plotList <- list()
chnkVec <- as.vector(NULL, mode = "character")
library(Seurat)
library(destiny)
dfPCA <- OsC@meta.data
dfPCA <- dfPCA[,grep("PC_", names(dfPCA))]
dmPCA <- DiffusionMap(dfPCA)
#dpt <- DPT(dm, tips = 268)
#dpt <- DPT(dm)
dpt <- DPT(dmPCA)
#pseudotime <- dpt$dpt
# Plot DC1 vs DC2 and color the cells by their inferred diffusion pseudotime.
# We can accesss diffusion pseudotime via dpt$dpt.
df <- data.frame(
DC1 = eigenvectors(dmPCA)[, 1],
DC2 = eigenvectors(dmPCA)[, 2],
DC3 = eigenvectors(dmPCA)[, 3],
"DM_Pseudotime" = dpt$dpt
)
df$cellID <- row.names(dfPCA)
## For this project reverse pseudotime ##
if (invertPT){
PTmax <- max(df$DM_Pseudotime)
df$DM_Pseudotime <- -1* (df$DM_Pseudotime - PTmax)
## Invert DC1, 2, 3
df$DC1 <- -1* df$DC1
df$DC2 <- -1* df$DC2
df$DC3 <- -1* df$DC3
}
## Add to table ##
df$cellID <- row.names(dfPCA)
dfdbTable <- OsC@meta.data
dfdbTable$row_names <- NULL
dfdbTable$DC1 <- NULL
dfdbTable$DC3 <- NULL
dfdbTable$DC2 <- NULL
dfdbTable$DM_Pseudotime <- NULL
dfdbTable <- unique(dfdbTable)
###############################################################################
## Add pseudotime components ##
#dim(dfdbTable)
dfdbTable[["cellID"]] <- row.names(dfdbTable)
dfdbTable <- merge(
dfdbTable,
df,
by.x = "cellID",
by.y = "cellID"
)
#dim(dfdbTable)
## Add to Seurat object ##
OsC@meta.data <- dfdbTable
## Create Pseudotime plot ##
dfTemp <- dfdbTable
#dotsize <- 0.5
dotcolor <- "darkblue"
tag <- "PC1PC2all"
plotList[[tag]] <- ggplot(dfTemp, aes(PC_1, PC_2, color=DM_Pseudotime)
)+ geom_point(
shape = 16,
size = as.numeric(dotsize)
) + xlab("PC1") + ylab("PC2") + scale_color_gradient2(
low="#ff6600",
high=dotcolor #,
#limits=c(0,maxExpr)
) + theme_bw(
) + theme(
axis.text.y = element_text(size=8),
axis.text.x = element_text(size=8),
axis.title.y = element_text(size=8),
axis.title.x = element_text(size=8),
axis.line = element_line(colour = "black"),
panel.border = element_rect(colour = "black", fill=NA, size=1),
plot.title = element_text(hjust = 0.5, size = 12),
panel.background = element_rect(fill = "lightgrey")
) + ggtitle("PC1, PC2 and DM Pseudotime"
) #+ xlim(minX, maxX) + ylim(minY, maxY)
## Save to file ##
FNbase <- paste0("Pseudotime_overview", VersionPdfExt)
FN <- paste0(Obio@parameterList$reportFigDir, FNbase)
FNrel <- paste0("report_figures/", FNbase)
pdf(FN)
print(plotList[[tag]])
dev.off()
## Create R markdown chunk ##
figLegend <- paste0(
'**Figure ',
figureCount,
'**: Figure depicting PCA components 1 and 2 with the diffusion map pseudotime highlighted in color. Download a pdf of this figure <a href="',FNrel,'" target="_blank">here</a>. '
)
figureCount <- figureCount + 1
NewChnk <- paste0(
"\n#### Pseudotime All Timepoints",
"\n```{r ", tag, ", results='asis', echo=F, eval=TRUE, warning=FALSE, fig.cap='",
figLegend,"'}\n",
"\n",
"\n print(plotList[['",tag,"']])",
"\n cat( '\n')",
"\n\n\n```\n"
)
chnkVec <- c(
chnkVec,
NewChnk
)
} # end dmap
if (length(plotList) > 3){
tabVar <- ".tabset .tabset-fade .tabset-dropdown"
} else {
tabVar <- ".tabset .tabset-fade .tabset-pills"
}Pseudotime
###############################################################################
## Do category enrichment on clusters ##
if (Obio@parameterList$addDmaps){
cat(paste(knit(text = chnkVec, quiet = T), collapse = '\n'))
}Pseudotime All Timepoints
Figure 72: Figure depicting PCA components 1 and 2 with the diffusion map pseudotime highlighted in color. Download a pdf of this figure here.
## Done doing enrichment on clusters ##
################################################################################## Will save Obio object here, so it can be re-used with different parameters
save(Obio,
file = paste0(
Obio@parameterList$localWorkDir,
Obio@parameterList$project_id,
".bioLOGIC.Robj"
)
)
print("Obio Object saved.")
save(OsC,
file = paste0(
Obio@parameterList$localWorkDir,
Obio@parameterList$project_id,
".Seurat.Robj"
)
)## Try to retrieve project data from db ##
library(RMySQL)
db.pwd2 <- "_asf_"
db.user2 <- "asf"
host2 <- "ms1.thecrick.org"
projectParams <- Obio@documentationParams
tryCatch({
dbDB = dbConnect(drv = RMySQL::MySQL(), user = db.user2, password = db.pwd2, host = host2, dbname = "asf");
dfProposal = dbGetQuery(dbDB, paste0("SELECT * FROM asf_proposals WHERE project_name ='",Obio@parameterList$lims.id,"'"));
dbDisconnect(dbDB)
}, error = function(x) {
message("Project Database could not be reached or has no entry in Obio@parameterList$lims.id for this analysis.")
})[1] TRUE
if (nrow(dfProposal) == 1){
if (!is.na(dfProposal[1,"ProjectAlias"]) & dfProposal[1,"ProjectAlias"] != ""){
projectParams[["title"]] = paste0(dfProposal[1,"ProjectAlias"], " - ", dfProposal[1,"project_name"])
}
if (!is.na(dfProposal[1,"project_user"]) & dfProposal[1,"project_user"] != ""){
projectParams[["subtitle"]] = paste0(dfProposal[1,"user_lab"], " Lab - ", dfProposal[1,"project_user"])
projectParams[["subtitle"]] <- gsub("^ Lab - ", "", projectParams[["subtitle"]])
}
if (!is.na(dfProposal[1,"proposal_text"]) & dfProposal[1,"proposal_text"] != ""){
projectParams[["abstract"]] = dfProposal[1,"proposal_text"]
}
}
## Escape all special characters
projectParams <- lapply(
projectParams, function(x)
#gsub("([.|()\\^{}+$*?]|\\[|\\])", "\\\\\1", x)
gsub("([.|()/\\^{}+$*?]|\\[|\\])", " ", x)
)
projectParams <- lapply(
projectParams, function(x)
#gsub("([.|()\\^{}+$*?]|\\[|\\])", "\\\\\1", x)
gsub("\\\n", " ", x)
)
#projectParams$title <- "Title"
# projectParams$abstract <- "This is the QC section."
#projectParams$subtitle <- "Abstract"Documentation
sessionInfo()## R version 3.6.0 (2019-04-26)
## Platform: x86_64-pc-linux-gnu (64-bit)
## Running under: CentOS Linux 7 (Core)
##
## Matrix products: default
## BLAS: /camp/apps/misc/stp/babs/manual/software/r/R-3.6.0-foss-2016b/lib64/R/lib/libRblas.so
## LAPACK: /camp/apps/misc/stp/babs/manual/software/r/R-3.6.0-foss-2016b/lib64/R/lib/libRlapack.so
##
## locale:
## [1] LC_CTYPE=en_GB.utf-8 LC_NUMERIC=C
## [3] LC_TIME=en_GB.utf-8 LC_COLLATE=en_GB.utf-8
## [5] LC_MONETARY=en_GB.utf-8 LC_MESSAGES=en_GB.utf-8
## [7] LC_PAPER=en_GB.utf-8 LC_NAME=C
## [9] LC_ADDRESS=C LC_TELEPHONE=C
## [11] LC_MEASUREMENT=en_GB.utf-8 LC_IDENTIFICATION=C
##
## attached base packages:
## [1] parallel stats4 stats graphics grDevices utils datasets
## [8] methods base
##
## other attached packages:
## [1] destiny_2.14.0 enrichR_2.1
## [3] GSEABase_1.46.0 graph_1.62.0
## [5] annotate_1.62.0 XML_3.99-0.3
## [7] AnnotationDbi_1.46.1 clusterProfiler_3.12.0
## [9] DT_0.14 AUCell_1.6.1
## [11] sleepwalk_0.3.0 ggtree_1.16.6
## [13] scales_1.1.1 ggrepel_0.8.2
## [15] RMySQL_0.10.20 DBI_1.1.0
## [17] DESeq2_1.24.0 SummarizedExperiment_1.14.1
## [19] DelayedArray_0.10.0 BiocParallel_1.18.1
## [21] matrixStats_0.56.0 Biobase_2.44.0
## [23] GenomicRanges_1.36.1 GenomeInfoDb_1.20.0
## [25] IRanges_2.18.3 S4Vectors_0.22.1
## [27] BiocGenerics_0.30.0 knitr_1.29
## [29] forcats_0.5.0 stringr_1.4.0
## [31] purrr_0.3.4 readr_1.3.1
## [33] tidyr_1.1.0 tibble_3.0.1
## [35] tidyverse_1.3.0 ggplot2_3.3.2
## [37] Seurat_3.1.5 dplyr_1.0.0
##
## loaded via a namespace (and not attached):
## [1] rappdirs_0.3.1 ggthemes_4.2.0 R.methodsS3_1.8.0
## [4] jrc_0.3.1 acepack_1.4.1 bit64_0.9-7
## [7] irlba_2.3.3 R.utils_2.9.2 data.table_1.12.8
## [10] rpart_4.1-15 RCurl_1.98-1.2 generics_0.0.2
## [13] cowplot_1.0.0 RSQLite_2.2.0 RANN_2.6.1
## [16] europepmc_0.4 proxy_0.4-24 future_1.17.0
## [19] bit_1.1-15.2 enrichplot_1.4.0 xml2_1.3.2
## [22] lubridate_1.7.9 httpuv_1.5.4 assertthat_0.2.1
## [25] viridis_0.5.1 xfun_0.15 hms_0.5.3
## [28] evaluate_0.14 promises_1.1.1 DEoptimR_1.0-8
## [31] fansi_0.4.1 progress_1.2.2 dbplyr_1.4.4
## [34] readxl_1.3.1 igraph_1.2.5 geneplotter_1.62.0
## [37] htmlwidgets_1.5.1 ellipsis_0.3.1 RSpectra_0.16-0
## [40] crosstalk_1.1.0.1 backports_1.1.8 vctrs_0.3.1
## [43] TTR_0.23-6 ROCR_1.0-11 abind_1.4-5
## [46] RcppEigen_0.3.3.7.0 withr_2.2.0 ggforce_0.3.2
## [49] triebeard_0.3.0 robustbase_0.93-6 vcd_1.4-7
## [52] checkmate_2.0.0 sctransform_0.2.1 treeio_1.8.2
## [55] xts_0.12-0 prettyunits_1.1.1 cluster_2.0.8
## [58] DOSE_3.10.2 ape_5.4 segmented_1.2-0
## [61] lazyeval_0.2.2 laeken_0.5.1 crayon_1.3.4
## [64] genefilter_1.66.0 pkgconfig_2.0.3 labeling_0.3
## [67] tweenr_1.0.1 nlme_3.1-139 nnet_7.3-12
## [70] rlang_0.4.6 globals_0.12.5 lifecycle_0.2.0
## [73] modelr_0.1.8 rsvd_1.0.3 cellranger_1.1.0
## [76] polyclip_1.10-0 lmtest_0.9-37 Matrix_1.2-18
## [79] urltools_1.7.3 carData_3.0-4 boot_1.3-22
## [82] zoo_1.8-8 reprex_0.3.0 base64enc_0.1-3
## [85] ggridges_0.5.2 png_0.1-7 viridisLite_0.3.0
## [88] rjson_0.2.20 bitops_1.0-6 R.oo_1.23.0
## [91] KernSmooth_2.23-15 blob_1.2.1 qvalue_2.16.0
## [94] jpeg_0.1-8.1 gridGraphics_0.5-0 memoise_1.1.0
## [97] magrittr_1.5 plyr_1.8.6 ica_1.0-2
## [100] zlibbioc_1.30.0 compiler_3.6.0 RColorBrewer_1.1-2
## [103] fitdistrplus_1.1-1 cli_2.0.2 XVector_0.24.0
## [106] listenv_0.8.0 patchwork_1.0.1 pbapply_1.4-2
## [109] htmlTable_2.0.0 Formula_1.2-3 MASS_7.3-51.4
## [112] mgcv_1.8-28 tidyselect_1.1.0 stringi_1.4.6
## [115] highr_0.8 yaml_2.2.1 GOSemSim_2.10.0
## [118] locfit_1.5-9.4 latticeExtra_0.6-29 grid_3.6.0
## [121] fastmatch_1.1-0 tools_3.6.0 rio_0.5.16
## [124] future.apply_1.6.0 rstudioapi_0.11 foreign_0.8-71
## [127] gridExtra_2.3 smoother_1.1 scatterplot3d_0.3-41
## [130] farver_2.0.3 Rtsne_0.15 ggraph_2.0.3
## [133] digest_0.6.25 rvcheck_0.1.8 BiocManager_1.30.10
## [136] shiny_1.5.0 Rcpp_1.0.4.6 car_3.0-8
## [139] broom_0.5.6 later_1.1.0.1 RcppAnnoy_0.0.16
## [142] httr_1.4.1 kernlab_0.9-29 colorspace_1.4-1
## [145] ranger_0.12.1 rvest_0.3.5 fs_1.4.2
## [148] reticulate_1.16 splines_3.6.0 uwot_0.1.8
## [151] tidytree_0.3.3 graphlayouts_0.7.0 sp_1.4-2
## [154] ggplotify_0.0.5 plotly_4.9.2.1 xtable_1.8-4
## [157] jsonlite_1.7.0 tidygraph_1.2.0 UpSetR_1.4.0
## [160] R6_2.4.1 Hmisc_4.4-0 pillar_1.4.4
## [163] htmltools_0.5.0 mime_0.9 VIM_6.0.0
## [166] glue_1.4.1 fastmap_1.0.1 class_7.3-15
## [169] codetools_0.2-16 fgsea_1.15.0 tsne_0.1-3
## [172] lattice_0.20-38 mixtools_1.2.0 curl_4.3
## [175] leiden_0.3.3 zip_2.0.4 openxlsx_4.1.5
## [178] GO.db_3.8.2 survival_3.2-3 rmarkdown_2.3
## [181] munsell_0.5.0 e1071_1.7-2 DO.db_2.9
## [184] GenomeInfoDbData_1.2.1 haven_2.3.1 reshape2_1.4.4
## [187] gtable_0.3.0The Francis Crick Institute, stefan.boeing@crick.ac.uk↩